Leveraging RNAs to Improve Prediction of Cancer Diagnosis and Prognosis

Abstract

The rapid development of high-throughput biotechnologies and analytical methodologies have brought new dimensions to cancer diagnosis and treatment, transforming traditional medicine towards precision medicine. Different from traditional health care adopting one-size-fits-all treatment, precision medicine is a much more advanced paradigm aiming to provide the right therapy to the right candidate patient at the right time. Transcriptomics, as crucial tumor biomarkers, have gained novel insights with their high-quality, cost-efficient, easily accessible for the accurate, robust, and efficient diagnosis, prognosis, disease monitoring, and prediction of drug response.

Ribonucleic acid (RNA) transcripts could be broadly divided into two types, protein-coding messenger RNAs (mRNAs) and non-protein-coding RNAs (ncRNAs), which could be further classified into microRNAs (miRNAs), piwi-interacting RNAs (piRNAs), circular RNAs (circRNAs), small interfering RNAs (siRNAs), etc. As an essential part of central dogma, mRNAs firstly stepped into tumor biomarker field, for their characteristics like delivering genetic and regulatory information and reflecting cellular states. Then ncRNAs followed to be noticed for thier main role in regulating mRNAs, stability and possible higher tissue-specificity. Although commonly used biomarkers such as prostate-specific antigen (PSA) in prostate cancer (PCa), alpha-fetoprotein (AFP) in hepatocellular carcinoma (HCC) were applied to clinical management for decades, their sensitivity and specificity still left a large room for improvement.

The work in this thesis focuses on RNA-based biomarkers to improve prediction of cancer diagnosis and prognosis, especially in PC and HCC. The study covers from the commonly used transcriptomics biomarkers like mRNA and miRNAs, to novel RNA molecules such as piRNAs and circRNAs. Using bioinformatics skills and machine learning to select candidate molecules, construct models and then modify panels in clinical cohorts. The aim is to improve the biomarker performance in both cancer diagnosis and prognosis. The studies in each chapter are described below:

Chapter 1: An introduction of RNA-based biomarkers and their clinical application. We distilled an integrated workflow encompassing important steps for RNA biomarker development.

Chapter 2: This chapter introduces an RNA-based biomarker, focusing on m6A RNA methylation-based gene signature, which was demonstrated to play a critical role in HCC initiation and development. We identified a gene signature panel and verified its prognostic performance in clinical cohorts, exhibiting a significant improvement in predicting overall survival (OS) in HCC patients.

Chapter 3: In this chapter, we further explore the potential of miRNAs as HCC microvascular invasion (MVI) diagnosis biomarker. We used comprehensive expression profiling of MVI-positive and negative serum specimens to identify an optimized transcriptomic panel of four miRNAs. After modifying the risk-stratification model combined with clinical pathological factors, which is dramatically superior to current clinical features.

Chapter 4: This chapter we describe the prospective validation of a newly developed serum miRNA panel for detection of clinically significant prostate cancer (PCa) in Chinese men. Over-investigation with invasive prostate biopsy, and subsequent overdiagnosis and overtreatment of indolent disease are inevitable for current PCa clinical management. This research may potentially develop into a new miRNA-based diagnostic test for prostate cancer, with the aim to accurately detect high-grade csPCa while avoiding over-diagnosing low-grade PCa.

Chapter 5: Exploration of the potential of novel RNAs like piRNAs and circRNAs as tumor prediction biomarkers. Two cases are introduced. We used piRNAs to monitor the development of castration resistant prostate cancer (CRPC), which is the outcome of the majority of PCa patients under deprivation therapy. Moreover, we tried the circRNAs analysis for HCC tissue specimen, suggesting competing endogenous RNAs (ceRNAs) may could provide a new dimension of understanding oncogene regulation.

Chapter 6: Summary of the major findings in the thesis and discussion were given.

Date of Award	23 Sept 2022
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	M YANG (Supervisor)