Circulating secretoneurin concentrations predict mortality and heart failure hospitalizations but not ventricular arrhythmias in stable ICD outpatients
H Rosjo, P L Myhre, E Riveland, A H Ottesen, N Sourour, H Kjekshus, A I Larsen, T OmlandAbstract
Background
Secretoneurin (SN) is a biomarker that provides strong prognostic information in patients with acute and chronic heart failure (HF). Whether SN concentrations predict mortality in stable outpatients with implantable cardioverter-defibrillators (ICDs), and whether potential increased risk relates to HF worsening or ventricular arrhythmias, are not known.
Purpose
To examine whether circulating SN concentrations can predict incident ventricular arrhythmias (primary endpoint) and mortality and HF hospitalizations (secondary endpoints) in high-risk outpatients with ICD.
Methods
We included 489 stable patients treated with ICD for both primary and secondary prevention from the outpatient service of two academic centers in Norway. SN concentrations were measured with prototype research-use only automated SN chemiluminescence immunoassay. We defined ventricular arrhythmia as either ventricular fibrillation or sustained ventricular tachycardia >100 b.p.m for minimum 30 sec, or appropriately delivered ICD therapies, i.e. electrical shock or anti-tachycardia pacing. Information regarding HF hospitalizations and mortality was obtained by review of electronic medical records.
Results
Median SN concentration was 33 pmol/L (quartile 1-3 26-41 pmol/L) and patients with high SN concentrations had higher prevalence of several risk factors, lower body mass index (BMI), and lower estimated glomerular filtration rate (eGFR) (Table). During mean follow-up 3.1±0.7 years, 136 patients (28%) experienced ventricular arrhythmia, 86 patients (18%) were hospitalized for HF, and 76 patients (16%) died. SN concentrations were associated with HF hospitalizations (hazard ratio [log] SN 4.6 [95% confidence interval 2.8-7.6], p<0.001) and mortality (hazard ratio 7.0 [4.2-11.5], p<0.001), but not with ventricular arrhythmia (hazard ratio 1.3 [0.8-2.1], p=0.3). The associations between SN quartiles and HF hospitalizations and mortality are presented in the Figure. Adjusting for age, sex, BMI, history of coronary artery disease and HF, and left ventricular ejection fraction (LVEF), SN concentrations (log transformed) were associated with HF hospitalization (hazard ratio 3.6 [1.8-7.3], p<0.001) and mortality (hazard ratio 4.7 [2.2-9.9], p<0.001). Adjusting also for eGFR, SN concentrations were still significantly associated with mortality and borderline significant for HF hospitalizations (p=0.07).
Conclusion
SN concentrations measured in stable ICD outpatients predict mortality and HF hospitalizations during follow-up, but not incident arrhythmic events. Accordingly, SN seems to provide strong prognostic information in stable outpatients with ICD, and more linked to HF worsening than to incident ventricular arrhythmia.