Identification of Image Phenotypes to Predict Recurrence After Resection of Hepatocellular Carcinoma

Overview

Tumor recurrence, which occurs in 70% of patients with HCC within 5 years after hepatic resection, is a major cause of post-resection-death. This recurrence can be true recurrence (intrahepatic metastases), which occurs sooner than 2 years later, or it can be due to the development of de-novo tumors at least 2 years later. Despite this high rate of tumor recurrence, no anti-recurrence adjuvant therapies are currently recommended. Imaging phenomics is the systematic, large scale extraction of imaging features for the characterization and classification of disease phenotypes. Combining imaging and tissue phenomics could be a solution to predict HCC recurrence. With the emergence of molecular therapies and immunotherapies, identifying patients with HCC at high risk of post-resection recurrence would help determine additional therapeutic and management strategies in clinical practice.

Hepatocellular carcinoma (HCC) is among the most lethal and prevalent cancers in the human population and it is now the third leading cause of cancer deaths worldwide, with over 500,000 people affected. Because of the high recurrence rate after curative hepatectomy, accurate prognostic assessment in HCC patients are quite important. With the emergence of molecular therapies and immunotherapies, the identification of patients at high or low risk for recurrence after hepatic resection would help determine additional therapeutic and management strategies in clinical practice. Although many immunohistochemical markers have been reported to have a prognostic value for HCC patients, there is no consensus on how these markers could add prognostic value to the clinical parameters. In the initial step of biomarker discovery, no specific sample size is provided, however to test hypothesis, 100 patients are required. This first study will potentially be followed by a second similar study promoted by the same investigators to increase the statistical power to improve the classification tool according to the patient's future. Period covered by the data collection: 2011-2019 / Duration data collection: 1 year. The primary endpoint will be built using machine learning method to obtain prediction of recurrence within 2 years. The Recurrence Free survival (RFS) within two years will be the reference outcome to evaluate the prognostic of the patients. The secondary endpoint are following : \- A secondary endpoint which will be built using machine learning method to obtain prediction of recurrence after 2 years. The Recurrence Free survival (RFS) after two years will be the reference outcome to evaluate the prognostic of the patients. \- A secondary endpoint will be the correlation between biomarker from CT scan and pathological biomarkers As the spectrum of HCC disease is very large, many patients to conduct conclusive validation studies for diagnostic and prognostic relevance need to be obtained. Overall, each specific-read out endpoint will include a sample size calculation and - if appropriate - a power analysis specific to the objective of this study. During training, phenotyping system performance assessment will be done to guide the calculation of the sample size for the validation.