Plasma microRNA Profiling Reveals Novel Biomarkers of Epicardial Adipose Tissue: A Multidetector Computed Tomography Study

Figures

• 

  Figure 1. Plasma microRNA (miRNA) profiling. (A) Unsupervised hierarchical clustering. The heat map diagram shows the result of a two-way hierarchical clustering of patients and miRNAs. Each column represents a patient (Tertile 1 of epicardial fat volume vs. Tertile 3 of epicardial fat volume). Each row represents a miRNA. The patient clustering tree is shown on top. The miRNA clustering tree is shown on the left. The color scale illustrates the relative expression level of miRNAs. The expression intensity of each miRNA in each sample varies from red to blue, which indicates relatively high or low expression, respectively. (B) p value for the comparison between study groups. Each point represents a miRNA. Red dots represent the selected candidates. (C) Plasma expression levels of miRNAs in study groups. (D) Expression levels of miRNAs in epicardial adipose tissue explants. Each point represents a sample. (E) Expression levels of miRNAs in conditioned media exposed to epicardial adipose tissue explants. Each point represents a sample. Relative quantification was performed using cel-miR-39-3p as the external standard for extracellular miRNAs and SNORD48 as the internal standard for tissue…

• 

  Figure 2. Plasma microRNA (miRNA) validation. (A–B) Examples of multidetector computed tomography scans and the corresponding epicardial fat volume of patients in the first-second and third tertiles of epicardial fat volume. (C) Plasma expression levels of miRNAs in study groups. MicroRNA levels were log2-transformed. MicroRNA expression levels are expressed as arbitrary units. Differences between groups were analyzed using Student’s t-test for independent samples. p values describe the significance level of differences for each comparison. EFV: epicardial fat volume.

• 

  Figure 3. Plasma microRNAs (miRNAs) as biomarkers of epicardial fat volume, according to EFV tertiles. (A) Area under the ROC curve (AUC) for each individual miRNAs and for combinations of miRNAs in pairs or families. (B) Performance of plasma miRNAs as biomarkers. (C,D) Decision trees calculated by chi-squared automatic interaction detector (CHAID) algorithm. The following variables were included in the clinical model: age, sex, body mass index and diabetes mellitus. MicroRNA levels were log2-transformed. For logistic regression models, data are presented as an odds ratio (OR) and 95% confidence intervals (CI). For discrimination analysis, data are presented as the AUC and 95% CI. For reclassification analysis, data are presented as the Integrated Discrimination Improvement (IDI) index and Net Reclassification Improvement (NRI) index and their respective and 95% CI. For decision trees, data are shown as frequency (percentage) of patients in each study group.

• 

  Figure 4. Plasma microRNAs (miRNAs) as biomarkers of epicardial fat volume, according to the cutoff values proposed by Spearman et al. [15]. (A) Area under the ROC curve (AUC) for each individual miRNAs and for combinations of miRNAs in pairs or families. (B) Performance of plasma miRNAs as biomarkers. (C,D) Decision trees calculated by Chi-squared Automatic Interaction Detector (CHAID) algorithm. The following variables were included in the clinical model: age, sex, body mass index and diabetes mellitus. MicroRNA levels were log2-transformed. For logistic regression models, data are presented as an odds ratio (OR) and 95% confidence intervals (CI). For discrimination analysis, data are presented as the AUC and 95% CI. For reclassification analysis, data are presented as the Integrated Discrimination Improvement (IDI) index and Net Reclassification Improvement (NRI) index and their respective 95% CI. For decision trees, data are shown as frequency (percentage) of patients in each study group.