#3101 LOW CIRCULATING LEVELS OF CARDIO-REGULATORY MIRNAS PREDICT WORSE OUTCOMES IN HEMODIALYSIS PATIENTS

Abstract

Abstract Background and Aims Patients with end stage renal disease (ESKD) on hemodialysis (HD) have a high risk of cardiovascular (CV) events, this being mostly related to the presence of uraemic cardiomyopathy. A deregulation in the expression of various microRNAs (miRNAs: non-coding RNAs that regulate post-transcriptional gene expression) has previously been implied in pathological cardiac remodelling. Nevertheless, scarce evidence has accrued so far on a possible significance in CV morbidity and mortality in HD patients, in relationship with the presence of uremic cardiomyopathy. Method We run a pilot, prospective, multicentre cohort study involving 74 ESKD patients undergoing chronic HD from 3 different hospitals in Italy and Greece. HD patients underwent a thorough clinical, laboratory and echocardiography assessment and were then prospectively followed for 24 months or until the occurrence of a composite endpoint of (CV and all-cause) mortality or non-fatal CV events. Through a systematic review of the literature, we identified a small panel of miRNAs (30a-5p, 23a-3p,451a and let7d-5p), which levels are known to be altered in either major CV disorders and kidney failure. miRNAs were then measured in the blood of HD patients and in a small group of matched healthy controls. Results miRNAs 23a-3p (p<0.0001), 451a (p = 0.001), 30a-5p (p = 0.003) and let7d-5p (p <0.0001) were all reduced in HD patients as compared with controls. Significant correlations were found between miRNAs and indexes of cardiac dysfunction such as Vmax, TAPSE and E/E’, as well as with some other laboratory parameters such as uric acid, sodium, potassium, HDL and CRP. During follow-up, 30/74 patients (40.5%) reached the composite endpoint. In these individuals, all miRNAs but let7d-5p were significantly reduced at baseline (p<0.0001). As showed by multivariate Cox-regression analyses, miRNAs 23a-3p, 451a, 30a-5p were all predictors of the composite endpoint (HR ranging from 0.943 to 0.995; p = 0.05 to 0.02) and Kaplan-Meier analyses confirmed a faster progression to the endpoint in subjects with miRNAs levels below an optimal ROC-derived cut-off value (p ranging from 0.001 to <0.0001; crude HRs 7.95 to 8.61; Figure 1). Conclusion Our study demonstrated that the evaluation of a very small panel of circulating miRNAs (30-5p, 23a–3p and 451a) may impart important prognostic information in chronic HD patients with respect to mortality and CV risk. Although the biological meaning of their deregulation remains unanswered, these preliminary findings may set the stage for larger investigation to generalize their usefulness as biomarkers, as well as possible therapeutic targets.