The overall goal of the project is to use mass spectrometry, nuclear magnetic resonance (NMR) analysis and large-scale non-coding gene expression analysis to search for new biomarkers characterize the transition steps of patients with chronic hepatitis or cirrhosis to liver cancer, thereby providing an appropriate mathematical model to support early diagnosis and screening of liver cancer in subjects with hepatitis B virus. Therefore, the specific goals of the project are:
- Establishment of a panel of biomarkers (including metabolism-related molecules and non-coding miRNAs) with high sensitivity to distinguish different clinical forms of chronic liver diseases related to HBV
- Optimizing the mathematical model based on newly discovered biomarkers for the purpose of screening and early detection of primary liver cancer in patients with chronic HBV infection.
Main tasks of the project
The research project will be divided into 2 phases as follows
- Phase 1 (discovery phase): At this stage, the technical process of NMR spectroscopy analysis is used to analyze metabolic markers. A high-throughput quantitative RT-PCR method will be used to detect changes in metabolic profiles and abnormal miRNA expression levels in the tested patients. The expected outcome of the discovery phase will be a novel but optimized biomarker panel consisting of metabolic molecules and miRNAs whose diagnostic performance will be confirmed in a later validation phase.
- Phase 2 (Validation and Standardization Phase): All testing and analytical parameters from the discovery phase will be applied to the validation phase and standardization of the accuracy capabilities of the diagnostic method based on biomarkers selected from the discovery phase. We plan to collect and analyze 600 patients (300 HCC, 100 LC, 200 CHB) using HPLC (and/or LCMS/MS) and high-throughput qRT-PCR methods. The raw data generated will be further analyzed using an analytical model (sPLS-DA); appropriate machine learning algorithms (random forests, linear support vector machines, PLS-DA and logistic regression) will be applied to build predictive models to monitor the progression of associated liver disease from HBV to HCC and helps support early diagnosis of liver cancer.
Project impact
- Diagnosis and treatment of cancer in general, including liver cancer, requires multi-modality in which searching for genetic lesions from peripheral blood is only a very small scope and certainly cannot completely solve the problem. Therefore the diagnosis of liver cancer, especially the early diagnosis of liver cancer, requires a combination of many different methods: clinical, imaging, biomarkers, molecular biology… In this project, we will establish prediction models based on metabolic markers and/or expression of miRNAs in liquid biopsy samples to help distinguish different disease stages in people with chronic HBV infection (chronic hepatitis, cirrhosis, liver cancer).
- Implications in practice: new and meaningful biomarkers to distinguish different disease stages in people with chronic HBV infection and to support early diagnosis of hepatocellular carcinoma in patients with chronic HBV infection is currently an urgent medical need. In this project, we established a panel of metabolic biomarkers and/or expression of miRNAs in liquid biopsy samples that distinguish different disease stages in chronic HBV-infected individuals with high sensitivity for disease progression and more importantly, new biomarker-based models can accurately/accurately detect early liver cancer. Combining research methods on changes in metabolic markers with expression levels of miRNAs provides a potential new tool for diagnosis, prognosis, monitoring and treatment decisions in clinical practice.