Unveiling cancer signals in liquid biopsy cell-free RNA profiles

Abstract

Liquid biopsies can be obtained in a minimally invasive way and as such form an interesting resource to look for cancer biomarkers. My thesis focuses on the underexplored potential of cell-free RNA (cfRNA). As an introduction to cfRNA biomarker identification, we performed a systematic review of cfRNA biomarker candidates for early ovarian cancer detection. While the majority of cfRNA studies focus on small RNAs, especially microRNAs, they are vastly outnumbered by messenger RNAs (mRNAs) and other long RNAs. Using advancements in mRNA capture sequencing, we analyzed and compared blood plasma mRNA profiles from cancer patients and healthy controls in a cohort covering 25 different cancer types. We also developed a novel strategy to leverage inherent variability for biomarker purposes. Non-coding RNAs have some interesting characteristics for liquid biopsy biomarker purposes as well. For circular RNAs, the tissue-specificity and stability in biofluids form assets for biomarker purposes. However, circular RNA detection in short-read RNA-sequencing is merely based on reads covering their characteristic back-splice junction. We developed a computational tool to better understand circular RNA contributions and assess potential biases in short-read RNA-sequencing read counts. Long non-coding RNAs (lncRNAs) may benefit from their cancer/tissue-specific expression for biomarker purposes, yet their low abundance complicates robust detection in biofluids. We therefore tested an in-house developed custom lncRNA probe panel to enrich for lncRNAs in tissue and biofluid samples. In summary, the field of cfRNA in cancer research is relatively young and fraught with challenges. My thesis contributes to the understanding of cell-free RNA profiles, but also emphasizes the need for standardized procedures, sensitive detection techniques, and apt data analysis methods.