Abstract 4760: Discovery of novel biomarkers against refractory cancer using in situ level mass spectrometry informatics

Abstract

Abstract Background and Purpose: The five-year survival rate of pancreatic cancer (PC) and diffuse-type gastric cancer (DT-GC) is far below 10%. They are known to have dense stroma unlike other types of tumors. Therefore, in the case, these two types of tumor are important to understand biological features of tumor stroma as well as that of tumor cells. However, conventional omics studies have largely focused on tumor cells. There are fewer studies in which both tumor cells and tumor stroma were examined simultaneously. In addition, among omics analysis, although genome and gene expression analysis using next generation sequencing (NGS) combined with informatics have progressed extensively, metabolomics analysis using mass spectrometry seems to be stagnated with no benefits from informatics. Hence, we have decided to develop in situ level mass spectrometry analysis combined with informatics in order to discover new biomarkers against both PC and DT-GC. Materials and Methods: First, we separated tumor cells, tumor stroma and normal cells using laser microdissection from clinical samples (5 cases of PC, 5 cases of DT-GC) and conducted trace-level metabolomics analysis using ultrasensitive capillary electrophoresis-mass spectrometry (CE-MS). We obtained mass spectrometric raw data with retention time, m/z and peak area of each molecules and then created data matrices. Using these data matrices, multivariate analysis such as PCA (principal component analysis) and PLS-DA (partial least squares regression analysis) was performed. Results: In PCA results, there were no significant differences among three data groups (tumor cells, tumor stroma and normal cells) and scatter plot with grouped data didn't follow any rules. Secondly, we conducted PLS-DA analysis and found that scatter plot was clearly divided into three groups. After calculating loading scores, we selected 20 top-ranked metabolites which is specific to each group (tumor cells, tumor stroma or normal cells) respectively. Moreover, all of them were unknown molecules and most of them had wide peak area and fast retention time. Now, we are identifying these molecules to find out which could be useful in the diagnosis and treatment of PC or DT-GC. Conclusion: We established in situ level mass spectrometry analysis and selected new biomarker candidates by loading plots in PLS-DA analysis. It is also indicated that informatics is a powerful approach to convert mass spectrometric raw data into meaningful information. Citation Format: Tamaki Hirakawa, Masahiro Yasunaga, Takayuki Kawai, Yoshihiro Shimizu, Kohei Shitara, Shigehiro Koganemaru, Yasutoshi Kuboki, Toshihiko Doi, Yasuhiro Matsumura. Discovery of novel biomarkers against refractory cancer using in situ level mass spectrometry informatics [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 4760.