Skip to main navigation Skip to search Skip to main content

Multiplex analysis of circRNAs in lung cancer using the nCounter technology

Student thesis: Doctoral thesis

Abstract

Lung cancer stands as one of the two most commonly diagnosed types of cancer, and the foremost cause of cancer-related death worldwide, primarily due to a late-stage diagnosis and lack of effective treatments for a significant number of patients. Although innovative single-agent therapies and their combination are constantly being tested in clinical trials, the five-year survival rate of late-stage lung cancer remains only at 5% (Cancer Research, UK). Consequently, advances in the early diagnosis of lung cancer are critical for improving the overall survival. Circular RNAs (circRNAs) are a re-discovered type of RNA showing a stable structure and tissue-specific expression. Accumulating evidence indicates how abnormal expression of some circRNAs can play a role in carcinogenesis and tumor progression, positioning these molecules as putative lung cancer biomarkers. In this context, many laboratories are currently investigating the clinical value of circRNAs found in liquid biopsies. Unfortunately, the lack of standardized methodologies for the study of circRNAs, together with some limitations inherent to liquid biopsies, is holding back their clinical implementation. The NanoString nCounter system can quantify the expression levels of up to 800 RNAs extracted from formalin-fixed paraffin-embedded (FFPE) material, cell lysates or liquid biopsies. Among the different advantages of this platform, we can highlight its short turnaround time and the ability to work with very low quantity of highly degraded samples. However, to the best of our knowledge, there is no published evidence on the use of this platform for the study of circRNAs in lung cancer specimens. Therefore, the main objective of this doctoral thesis was to test the nCounter technology for the study of circRNA expression in lung cancer samples, including cell lines, FFPE samples and liquid biopsies. In Chapter 2, we present a proof-of-concept study of nCounter for the detection of circRNAs in lung cancer samples. First, we designed and validated a customized nCounter circRNA panel in lung cancer cell lines. Then, the panel was used to study circRNA expression in tissue biopsies. As a result, we found a cluster of differentially expressed circRNAs in early and late-stage lung cancer (n=53) vs. non-cancer controls (n=16). Also, machine learning (ML) analysis allowed us to develop two circRNA signatures to discriminate early and late-stage lung cancer biopsies from non-tumor samples with high accuracy. In Chapter 3, we transitioned from solid specimens to liquid biopsies. In this second proof-of-concept study, we established and validated a protocol for the analysis of extracellular vesicle (EV)-circRNAs by nCounter. To this end, we tested different key points such as initial volume of plasma, EV purification method, number of cycles of pre-amplification, data normalization procedures or ML approaches. Then, we used the protocol to analyze the EV-circRNAs in plasma samples of lung cancer patients (n=36) and non-cancer controls (n=30). Again, we found a cluster of differentially expressed EV-circRNAs that could potentially be used as biomarkers of lung cancer. Also, ML allowed us to find a 10-circRNA signature that discriminated lung cancer samples from controls. Purification of EVs may result challenging to implement in the clinical setting. Consequently, in Chapter 4, we tested nCounter in whole plasma and detected a higher number of circRNAs compared to EV-enriched samples. Then, we analyzed plasma samples of early-stage lung cancer patients (n=49) and non-cancer controls (n=49), which included individuals with benign nodules (n=19/49). Subsequently, ML techniques generated a signature to discriminate early-stage lung cancer with an area under the ROC curve (AUC ROC) of 0.9. In conclusion, this thesis has proved that nCounter can be used for the study of circRNAs in lung cancer samples, setting the ground for the development of clinically relevant circRNA assays.
Date of Award23 Nov 2022
Original languageEnglish
SupervisorRafael Rosell Costa (Director)

Cite this

'