Differential mortality risk prediction by aldosterone and cortisol in systolic heart failure with versus without mineralocorticoid receptor blockade

Abstract

Background: We have previously shown that high levels of aldosterone and cortisol predict an adverse prognosis in patients with systolic and non- systolic heart failure. Mineralocorticoid receptor antagonists (MRA) improve survival and increase levels of both these steroid hormones. However, the prognostic significance of serum aldosterone and cortisol levels in patients treated with MRA has so far remained unclear. Methods: We used data of the randomized Extended Interdisciplinary Network for Heart Failure (E-INH) trial which investigates effects of nurse-coordinated multidisciplinary disease management in patients with systolic heart failure (left ventricular ejection fraction [LVEF] ≤40%). Baseline cortisol and aldosterone could be measured in 903 patients. Median follow-up was 37 months (quartiles 27, 43). Results: Mean age of the study cohort was 68±12 y, 28% female, mean EF 31±8%, 45% in NYHA functional class III or IV, 43% on MRA. During follow-up 328 patients died (36%). There was a trend towards a lower crude mortality risk in patients with vs without MRA (33% vs 39%, P=0.069). Aldosterone but not cortisol independently predicted increased mortality risk in Cox regression analyses after adjustment for age, sex, NYHA class, EF, NT-proBNP, renal function, C-reactive protein, sodium, and hypercholesterolemia: hazard ratio (HR) for highest vs lowest tertile of aldosterone 1.58 (95% CI 1.20-2.01; P=0.001). We then performed stratified analyses in patients with or without MRA. In patients without MRA, higher levels of both aldosterone and cortisol predicted an increased mortality risk within the same multivariable model: HR for highest vs lowest tertile for aldosterone 1.47 (1.03-2.11; P=0.036), and for cortisol 1.68 (1.15-2.43; P=0.007), respectively. In patients with MRA, aldosterone predicted risk similarly: HR for highest vs lowest tertile 1.99 (1.24-3.17; P=0.004). By contrast, the prognostic effect of cortisol pointed in the opposite direction: HR for highest vs lowest tertile 0.62 (0.39-0.98; P=0.04). Conclusions: In patients with systolic heart failure, MRA appear to differentially influence the pathophysiologic mechanisms, and hence prognosis, exerted by aldosterone and cortisol. MRA are known to increase both, aldosterone and cortisol levels. In subjects on MRA, high cortisol predicts a lower mortality risk, thus possibly indicating a better response to MRA treatment.