Impact of circulating exosomal long non-coding RNA H19 on clinical outcomes in patients with heart failure
T Koya, T Nagai, M Nakao, Y Mizuguchi, A Takahashi, Y Kato, S Imagawa, Y Tokuda, M Takahashi, J Matsumoto, K Motoi, Y Wakamatsu, M Machida, T Saito, T AnzaiAbstract
Background
Although long non-coding RNA H19 suppresses cardiac hypertrophy through both microRNA-675 (miR-675)–mediated and direct molecular mechanisms in experimental models, small cohort studies reported conflicting associations between circulating H19 levels and prognosis across different cardiovascular conditions.1-3 It remains unclear whether circulating H19 levels are associated with clinical outcomes in patients with heart failure (HF).
Purpose
To investigate the association between circulating H19 levels and clinical outcomes in patients with HF.
Methods
We prospectively examined 564 consecutive patients with chronic HF from a multicentre registry (29 sites) from January 2020 to December 2023. Circulating serum H19 expression was quantified using real-time polymerase chain reaction after RNA extraction, normalized to an internal control, and expressed using the 2−ΔCt method. Patients were stratified into lower and higher H19 groups according to the median ΔCt value (14.3). Patients were further categorized into four groups according to median H19 and NT-proBNP levels. The primary outcome was a composite outcome of all-cause death and HF hospitalisation.
Results
The study population comprised 322 men, with a mean age of 70 ± 14 years and a median LVEF of 45 (IQR 32–62)%. HFrEF was present in 237 patients (42.2%). Patients with higher H19 levels had higher LVEF and smaller left ventricular end-diastolic diameter, and left ventricular mass index compared with those with lower H19 levels. Circulating H19 levels were significantly correlated with miR-675 expression (r = 0.14, P < 0.001), but showed no significant association with log NT-proBNP (r = 0.001, P = 0.98) or hs-troponin T (r = 0.03, P = 0.52). During a median follow-up of 1547 (IQR 1050-1823) days, the primary outcome occurred in 97 patients. Higher H19 levels were significantly associated with a greater incidence of the primary outcome (Figure 1). When patients were categorized according to median H19 and NT-proBNP levels, those with both higher H19 and higher NT-proBNP levels exhibited the highest event rates (Figure 2). In a multivariable Cox regression analysis, higher H19 levels were independently associated with the primary outcome (HR 1.78, 95% CI 1.29-2.46) even after adjusting for after adjustment for age, sex, systolic blood pressure, NYHA class, serum sodium, eGFR, diabetes, LVEF, and log NT-proBNP. The association remained significant after additional adjustment for the MAGGIC risk score. Addition of H19 to a model including MAGGIC and NT-proBNP significantly improved model fit (ΔLR χ² = 8.44) and modestly increased discrimination (c-index 0.723 to 0.726, P = 0.004).
Conclusions
In patients with HF, higher circulating H19 levels were independently associated with adverse clinical outcomes. H19 provided incremental prognostic information beyond NT-proBNP and established risk score, supporting its clinical relevance for risk stratification in HF.