Association of soluble ST2 and right ventricular dysfunction with mortality in chronic hemodialysis patients

Abstract

Abstract Background End stage kidney disease (ESKD), is a triggering and facilitating factor for cardiac remodelling (i.e., morphologic hypertrophy and/or dilatation associated with deterioration of systolic and/or diastolic function) that contribute to heart failure (HF). RV dysfunction (RVD) has been demonstrated to predict mortality in ESKD patients. Purpose The present study aimed to investigate the potential associations between RVD and circulating biomarkers of myocardial inflammation and fibrosis with all-cause mortality in HD patients. Method We performed a retrospective single-centre cohort study of prevalent patients admitted in a chronic HD program for more than 3 months. Clinical characteristics and echocardiographic parameters were assessed in all patients. Pre-dialysis blood samples for measurement of inflammatory (e.g., C reactive protein, interleukin-1, interleukin-18) and fibrotic (e.g., soluble suppression of tumorigenesis-2 [sST2], galectin-3, C-terminal pro-peptide of procollagen type I and N-terminal pro-peptide of procollagen type III) biomarkers were collected. RVD was defined using tricuspid annular plane systolic excursion (TAPSE) <1.7 cm or pulsed Doppler peak annular velocity (S') <9.5 cm/s. The ability of sST2 to discriminate between mortality was assessed using AuROC curve. Results We enrolled forty-eight patients, mean patients age was 74 (64 – 79)years, and 62.5% were males. 95.8% of the patients had high blood pressure, and at least 70.8% had HF criteria. About 52.1% of the patients were on OnLine HDF. Regarding the echocardiogram parameters, 56.3% had no functional heart disease, 10.4% had LVD (LVEF ≤45% and diastolic dysfunction ≥ grade 2), and 33.3% had RVD (TAPSE <17mm and/or S'<9.5cm/s), with or without LVD. Mortality was higher 45.5% (log-rank, p=0.003) in patients with RVD as diagnosed by S' than in patients without RVD. No difference in mortality was observed for RVD defined by TAPSE. There were no differences in the morphology and function parameters of the left ventricle between patients with and without RVD. From all biomarkers measured only sST2 was associated with RVD. Indeed, an age- and sex-adjusted analyses showed that doubling of sST2 was inversely associated with a decreased in S' (estimate = −2.03, 95% CI [−3.04 to −1.00] cm/s; P=0.002). Mortality was increased in patients with sST2 ≥40.45 ng/mL compared to patients with sST2 <40.45 ng/mL (66.7% vs. 18.9%, log-rank; p=0.004). Conclusion This preliminary data would suggest that patients on chronic HD, circulating levels of sST2 were independently associated with RVD. In addition, elevated sST2 levels and RVD were associated with increased all-cause mortality. The myocardial pro-remodelling effect of sST2 in HD patients with RVD warrants further investigation. Funding Acknowledgement Type of funding sources: None.