Utility of miR-155, miR-21 and miR-16 in plasma as biomarkers of atrial fibrillation

Abstract

Abstract Background Atrial fibrillation (AF) is associated with increased morbidity and mortality. MicroRNAs (miRs) are post-transcriptional regulators that could be useful in the study of AF. Purpose To investigate whether miR-16, miR-21 and miR-155 may differentiate between AF and sinus rhythm. Methods 29 controls and 42 AF (June 2019-May 2020). Lipid profile, hsCRP were measured. Serum RNA extraction was performed. Hsa-miR-21, hsa-miR-16, hsa-miR-155, and cel-miR-39 were amplified RT-PCR. Quantitative results for miR levels were generated according to the 2-ΔCt method using cel-miR-39 for normalization. Correlations were assessed by Spearman rank test. Logistic regression analysis was adjusted by age, sex, and BMI vs miR's. A p<0.05 was set for significance. Results (controls vs AF): age (56 [52–63] vs 60 [64–67] years, p=0.122), BMI (30.52 [27.69–32.67] vs 29.65 [26.3–32.9] kg/m2 p=0.779), female 17 (58.6%) vs 19 (45.2%), p=0.268; total cholesterol (198.4 [17.2–215] vs 171.3 [150.3–216.8] mg/dL, p=0.066), LDL (126.6 [114.7–141.6] vs 99.2 [80.7–124] mg/dL, p=0.005) HDL (40.4 [36.8–45.8] vs 43.8 [36.9–48.5] p=0.326) and CRP (2.69 [1.22–4.72] vs 2.92 [1.4–4.76] mg/L p=0.835). The CHA2DS2VASC score in AF was <2=11 (45.2%) and ≥2= 26 (54.8%). MicroRNA levels (controls vs AF): miR-21 0.0195 (0.0095–0.0275) vs 0.042 (0.028–0.098) p<0.001; miR-155 0.0178 (0.0108–0.0309) vs 0.01 (0.0056–0.0169) p<0.001, and miR-16 0.5238 (0.2618–1.992) vs 0.1652 (0.0765–0.3816) p<0.001. In the logistic regression miR-21 had a positive correlation with AF, c=0.851 p=0.002, Wald=9.89. miR-16 and miR-155 a negative relation (c=0.76,0.72; p=0.036, 0.043, Wald=4.39, 4.085; respectively) (Fig. 1A, B, C). For multivariable analysis we used miR-16, miR-21, and miR-155 as previous analysis demonstrated significance with an overall model c=0.90 p=0.028 (Wald=14.2) and significance was seen in the miR-21 Wald= 9.33 (p=0.002); and miR-155 Wald=4.7 (p=0.03); miR-16 was not significant (Wald=2.7, p=0.1). A ternary plot was performed using the predicted outputs of the model. (Fig. 2) Conclusion miR-16, miR-21 and miR-155 are useful for the discrimination of AF vs sinus rhythm and may be part of important physiopathological pathways in AF. Funding Acknowledgement Type of funding sources: Public hospital(s). Main funding source(s): Instituto Nacional de Cardiología Ignacio Chávez Figure 1. Logistic regression curvesFigure 2. Multivariate ternary plot of miRs