Identification of clinically relevant profiles in Colorectal Cancer through integrated analysis of bacterial DNA and metabolome in serum

Juan Vicente-Valor¹Sofía Tesolato¹Dulcenombre Gomez-Garre²Mateo Paz- Cabezas²Adriana Ortega-Hernández²Constanza Fernández-Hernández³Sofía De La Serna²Inmaculada Domínguez-Serrano²Jana Dziakova²Daniel Rivera²Francisco Javier Rupérez³Antonia Garcia³Antonio Torres²Pilar Iniesta^1*

• ¹Department of Biochemistry and Molecular Biology and San Carlos Health Research Institute (IdISSC), Facultad de Farmacia, Universidad Complutense de Madrid, Madrid, Spain
• ²San Carlos Health Research Institute (IdISSC), San Carlos University Clinical Hospital, Madrid, Madrid, Spain
• ³Metabolomics and Bioanalysis Center (CEMBIO). Faculty of Pharmacy., CEU San Pablo University, Madrid, Madrid, Spain

There is increasing evidence demonstrating the relationship between microbiota and colorectal cancer. Several studies have been published analyzing microbiota in tissues and feces from cancer patients; however, there are only a few publications investigating the clinical utility of serum microbiome from colorectal cancer patients. Our aim was to advance in the search for serum biomarkers for the diagnosis of colorectal cancer. We conducted a cross-sectional study assessing bacterial DNA and metabolomic profiles in 64 serum samples from subjects affected by colorectal cancer and controls. A metagenomic analysis of the bacterial 16S rRNA gene in serum was established, and serum metabolites were detected through an untargeted metabolic study based on Gas Chromatography-Quadruple Time-Of-Flight Mass Spectrometry with accurate mass. After integrating the data resulting from the bioinformatics and statistical analyses, we obtained different profiles in colorectal cancer population and controls, regardless of the subjects' age, gender and body mass index. Serum levels of Firmicutes and threonic acid were the most relevant characteristics that could help differentiate both groups, achieving an excellent predictive accuracy in this discovery cohort (area under the ROC curve = 0.95). Although these results should be validated in other cohorts through multicenter studies, we consider that our data could be relevant and applicable to the early diagnosis of colorectal cancer.