Analysis of circulating tumor DNA by using NGS-based method with enhanced analytical performance

Abstract

Introduction: Interrogation of circulating tumor (ct)DNA using next-generation sequencing (NGS)-based methods have been proposed as a way to track the dynamics of tumor in real time. However, there was no standard guideline for ctDNA sequencing that I have evaluated the procedure from end-to-end to propose the optimal analysis methods for ctDNA sequencing. Chapter 1* emphasizes the importance of the recovery of unique DNA molecule from the minimal amount of starting material. After that, the systematic evaluation of each step highlights the error-prone step in the sequencing process. In Chapter 2, the utility of ctDNA sequencing has evaluated through the monitoring of tumor genomic in multi-cancer samples. Method: To maximize the recovery rate of unique DNA molecule, I approached the ligation step during the library preparation in sequencing protocol by optimizing the temperature, time and adapter concentration. Identification of technical errors was conducted with the comparison of background error distribution from the acoustically sheared germline DNA and naturally fragmented cell-free DNA. The utility of ctDNA sequencing analysis was assessed by comparing the standard protein biomarker and imaging changes during the patients therapeutic intervention. Results: The modified ligation conditions for the minimal amount of starting material able to increase the recovery rate of unique DNA molecule by 20% compared to the standard conditions. A comparison of the characteristic of acoustically sheared gDNA and naturally fragmented cfDNA revealed that gDNA constituted with 64% of C: G> A: T and 39% of C: G> G: C substitution class changes. Through testing of the series of the mild sheared conditions, the reduction of error rate was observed with an average of 40%. Furthermore, the analysis of the vicinity at the ends of the DNA fragments revealed that A> G and A> T preferentially fragmented. The enhanced analytical performance in NGS method able to establish diagnostic utility with the detection sensitivity of 100% and specificity of 97.1% as applied to cancer plasma samples. The level of ctDNA was not only highly correlated with the therapeutic response but also showed an average of two months earlier reaction than the standard protein biomarker and imaging changes. Finally, the determination of tumor heterogeneity was observed through ctDNA analysis, which was not discovered in the matched tumor biopsies. Conclusions: Overall, the unique characterization of cfDNA could not only emphasize the underlying cause of technical errors but also demonstrate opportunities for early detection of cancer using NGS-based technology. Ultimately, the combined approach of ctDNA and NGS sequencing analysis is believed to address unmet needs in cancer research.