MicroRNAs as disease specific diagnostic biomarkers for neoplastic aetiology-related and inflammatory-related pericardial fluid effusion

Abstract Background and aim Malignant involvement of the pericardium is seen in 1 to 20 percent of autopsies in patients with cancer. The most common metastatic tumor involving the pericardium is lung cancer [1]. We aimed to distinguish the origin of the pericardial fluid effusion (i.e. pericarditis vs cancer) based on miRNAs expression in peripheral blood plasma. Method 8 patients hospitalized for collection of pericardial fluid with pericardial effusion of neoplastic aetiology (lung cancer). Control group includes 8 patients with effusion of inflammatory aetiology. Plasma RNA was extracted by mirVANAPARISKit and quality of RNA was assessed by fluorometric assay. GEP analysis was performed using the Clariom D pico chips, analysed on the Affymetrix platform. Statistical analysis by TAC software. Additional analyses were performed in and R using Signal information obtained from the TAC output. We performed the following tests using log2 transformed data and all comparison groups (A-F). Additional FDR correction, logistic regression, Mann-whitney t-test was used depending of the variables. We calculated Area under the curve using ROCp R package. Scores were ranging from 0 to 1. Co-expression analysis to identify genes authentically expressed was performed using Spearman correlation (cutoff = 0.9, Rpval = 0.05). In order to identify the targets of DE miRNAs we used our wizbionet R package and previously developed pipelines [2,3]. We performed target screening using multimiR package, selecting top 20% predictions from all available databases. Results We analyzed targets for all mature versions, and if DE miR was identified as pre-miR we generated -3p and -5p version for it. We also screened DisgeNet database for genes associated with cancer and pericarditis, we identified 2823 and 157 such genes. After identification of the targets of DE miRNAs we performed data aggregation, summarization and obtained information how many targets overall and targets associated with IS are regulated by each DE miRNA. Additionally we identified top targets regulated by the top miRNAs. MiR-5695, miR-4446-5p, miR-572, miR-3131 and miR-4784 were found the most significantly differentially expressed miRNAs in blood plasma for patients with malignancy compared to pericarditis. MiR-22-3p, miR-642a, miR-6771, miR-140-3p, and miR660-5p were found the most significantly differentially expressed miRNAs in pericardial fluid plasma for patients with malignancy compared to pericarditis. Impo