Cardiovascular risk associated with serum potassium in the context of mineralocorticoid receptor antagonist use in patients with heart failure and left ventricular dysfunction

Patrick Rossignol*¹, Kevin Duarte¹, Nicolas Girerd¹, Moez Karoui¹, John J.V. McMurray², Karl Swedberg³, Dirk J. van Veldhuisen4, Stuart Pocock5, Kenneth Dickstein6, Faiez Zannad¹, Bertram Pitt7

¹ Université de Lorraine, Inserm, Centre d’Investigations Cliniques- Plurithématique 14-33, and Inserm
U1116, CHRU, F-CRIN INI-CRCT (Cardiovascular and Renal Clinical Trialists), Nancy, France
² The British Heart Foundation Cardiovascular Research Centre, University of Glasgow, Glasgow, UK
³ Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden
4 University of Groningen, University Medical Center Groningen, Groningen, The Netherlands
5 Department of Biostatistics, London School of Hygiene & Tropical Medicine, London, UK
6 Department of Cardiology, University of Bergen, Stavanger University Hospital, Stavanger, Norway
7 University of Michigan School of Medicine, Ann Arbor, MI, USA

Patrick Rossignol: ✉ 

*Correspondence to:

Corresponding author. Centre d’Investigations Cliniques-INSERM CHRU de Nancy, Institut lorrain du Coeur
et des Vaisseaux Louis Mathieu, 4 rue du Morvan, 54500 Vandoeuvre Lès Nancy, France. Tel: +33 3
83157320, Fax: +33 3 83157324, Email:



To assess the prognostic value of mineralocorticoid receptor antagonist (MRA) initiation and change in serum potassium (K+) during follow-up in patients post-acute myocardial infarction with left ventricular dysfunction or chronic heart failure (HF) and reduced ejection fraction (HFrEF).

Methods and results

Risk scores for predicting cardiovascular death (primary outcome), hospitalization for HF and all-cause death were developed. K+ and other relevant time-updated clinical and biological variables were added to conventional prognostic factors when constructing these new models. EPHESUS (n = 6632) was the derivation cohort, while EMPHASIS-HF (chronic HF, n = 2737) was used as external validation cohort. The final cardiovascular death risk score included medical history, clinical and biological parameters (e.g. K+, below or above the normal range of 4–5 mmol/L, estimated glomerular filtration rate, and anaemia), as well as aspects of treatment (any diuretic usage, MRA use or discontinuation, and beta-blocker use). The risk score performed well in both the derivation and validation cohorts and outperformed the MAGGIC score. A webbased calculator was created to allow easy determination of the risk score (


Adding time-updated variables, including K+ and MRA treatment, improved risk prediction of cardiovascular death (on top of the MAGGIC score) in patients with HF eligible for renin–angiotensin system inhibitors and MRA therapy. This new risk score including MRA usage and K+ may be of value in helping physicians to better use MRAs, avoid unnecessary and potentially detrimental permanent discontinuations, and therefore improving cardiovascular outcomes in patients with chronic HFrEF or HF after acute myocardial infarction with left ventricular dysfunction.

European Journal of Heart Failure in press

(Pubmed weblink : pending)