Correlation of the novel biomarker cellular communication network factor 1 (CCN1) with miR-155 in serum of patients with dilated cardiomyopathy

Abstract Background We recently reported an independent prognostic ability of serum CCN1 to predict all-cause mortality in DCM, irrespective of its aetiology (inflammatory DCM (DCMi) or conventional DCM). Interestingly, CCN1 levels were lower in patients with DCMi than in patients with conventional DCM. In a murine model, miR-155, which has key functions in inflammation, repressed the expression of CCN1 and impeded its function in angiogenesis. In a recent study, patients with DCMi had higher miR-155 plasma levels than those with conventional DCM. Purpose The present study evaluated whether circulating levels of CCN1 and miR-155 are associated in order to provide insight into the underlying pathophysiology. Methods Patients with a primary diagnosis of DCM, defined as LVEF <45% and an increased LVEDD (according to HENRY >117%), were included in this single-centre study. Exclusion criteria comprised primary valvular diseases (≥ second degree), acute myocarditis, active infectious diseases, pulmonary diseases, cancer, chronic alcoholism, and heart failure of other origins. CCN1 levels were determined in serum at study inclusion using an enzyme-linked immunosorbent assay. miR-155-5p was measured in serum using quantitative real-time PCR and normalized to a spike-in of miR-39-3p from C. elegans for analysis. The primary endpoint was all-cause mortality during follow-up. An adjusted multivariable Cox regression model (comprising age, sex, time of symptoms, LVEF, eGFR, CRP, and NT-proBNP) was used to assess the association between CCN1 or miR-155 with all-cause mortality. Results Among 283 patients in the DCM cohort with CCN1 data, miR-155 measurements were available for analysis in 270 patients. Among those, 77.4% were males at a median age of 55.3 [IQR 47.8, 64.7] years with 3.84 [IQR 1.1, 19.3] months since onset of disease who had severely reduced LVEF (31 [IQR 26, 37] %), elevated LVEDD (67 [IQR 62, 72] mm), and normal eGFR (CKD-EPI) (91.5 [IQR 74.01, 103.3] ml/min). During a median follow-up of 12.4 [IQR 10.5, 13.9] years, a total of 101 (37.4%) patients died. There was no significant difference in miR-155 levels between DCMi and conventional DCM (ΔCt: 14.84 vs 14.94, P = 0.60; and ΔΔCt values 1.68 vs 1.49, P = 0.42). A positive correlation was found for miR-155 (relative expression ΔΔCt values) with CCN1 in serum from patients with conventional DCM (rS = 0.251, p = 0.002) but not with DCMi (rS = 0.007, p = 0.94) (Figure 1). Unlike CCN1 (HR 1.11; 95% CI 1.04, 1.19; p = 0.002), miR-155 was not associated with all-cause mortality (HR 0.95; 95% CI 0.82, 1.11; P = 0.54) in adjusted multivariable Cox regression models. Conclusion A differential correlation of miR-155 with CCN1 levels was found in patients with conventional DCM vs. those with DCMi, suggesting a potential role of these biomarkers only in conventional DCM. Further investigation of the pathophysiology of CCN1 in DCM is warranted, including -omics analyses.