Incident Hyperkalemia, Hypokalemia, and Clinical Outcomes During Spironolactone Treatment of Heart Failure With Preserved Ejection Fraction: Analysis of the TOPCAT Trial

Akshay S. Desai, Jiankang Liu, Marc A. Pfeffer, Brian Claggett, Jerome Fleg, Eldrin F. Lewis, Sonja McKinlay, Eileen O'Meara, Sanjiv J. Shah, Nancy K Sweitzer, Scott Solomon, Bertram Pitt

Research output: Contribution to journalArticle

12 Scopus citations


Background: In patients with heart failure and preserved ejection fraction (HF-PEF) randomized in the Americas as part of the Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist (TOPCAT) trial, treatment with spironolactone enhanced the risk of hyperkalemia but reduced the risk of hypokalemia. We examined the clinical correlates and prognostic implications of incident hypo- and hyperkalemia during study follow-up. Methods: We defined the region-specific incidence of hypokalemia (potassium [K+] <3.5 mmol/l) and hyperkalemia (K+ ≥5.5 mmol/l) among both placebo- and spironolactone-assigned patients in TOPCAT. Factors associated with incident hypokalemia and hyperkalemia and the relationship between incident K+ abnormalities and the risk of subsequent mortality were analyzed in multivariable regression models restricted to the Americas. Results: In the Americas, assignment to spironolactone increased risk for hyperkalemia (hazard ratio 3.21, 95% confidence interval 2.46–4.20, P <.001) and reduced risk of hypokalemia (hazard ratio 0.43, 95% confidence interval 0.34–0.55, P <.001). Assignment to spironolactone, lower estimated glomerular filtration rate, higher baseline K+, diabetes, and lower hemoglobin were associated with incident hyperkalemia, whereas assignment to placebo, lower K+, younger age, lower estimated glomerular filtration rate, and use of diuretics at baseline were associated with hypokalemia. The combination of spironolactone and an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker was associated with incremental risk for hyperkalemia and protection from hypokalemia. Independent of region, both hypokalemia and hyperkalemia, were associated with higher risk for cardiovascular and all-cause mortality in multivariable-adjusted Cox regression models. Conclusions: Both hyperkalemia and hypokalemia are associated with heightened risk for mortality in HF-PEF. Use of spironolactone in this population requires careful laboratory surveillance of K+ and creatinine, particularly in high-risk groups.

Original languageEnglish (US)
Pages (from-to)313-320
Number of pages8
JournalJournal of Cardiac Failure
Issue number5
Publication statusPublished - May 1 2018



  • aldosterone
  • diastolic heart failure
  • heart failure
  • Hyperkalemia
  • hypokalemia
  • mineralocorticoid receptor
  • spironolactone

ASJC Scopus subject areas

  • Cardiology and Cardiovascular Medicine

Cite this