Background: Aldosterone antagonists (AAs) improve mortality in patients with chronic heart failure (CHF). However, they may also cause acute renal failure and hyperkalaemia. The relative risk-benefit ratio of the available AAs spironolactone, eplerenone and canrenone with respect to acute renal failure and hyperkalaemia is unclear.

Methods: We conducted a systematic review and network meta-analysis following PRISMA-P and PRISMA-NMA guidelines. 7 individual databases, 6 individual clinical trial registries, and 3 individual grey literature databases were searched up to January 2017 for randomized controlled trials with an active treatment of either spironolactone, eplerenone, or canrenone/potassium-canreonate in adults with symptomatic CHF due to systolic dysfunction (left ventricular ejection fraction < 40%) reporting on acute renal failure or hyperkalaemia. The pooled effect estimates were ranked by the surface under the cumulative ranking curve (SUCRA).

Results: We identified 5 trials including 4,923 CHF patients informing on acute kidney failure and hyperkaliaemia as safety endpoints. The relevant network plot is shown in figure 1.


Figure 1: Network plot. The coloured edges (green) indicate an average low-level of risk of bias in trials according to the GRADE) approach weighted according to the number of studies per comparison.


Beta-blocker adjusted predictive interval plots and cumulative ranking probabilities for SUCRA-values are shown in figures 2 -5.

They appear to favour spironolactone over eplerenone (OR: 3.47 [1.44-8.37]; spironolactone as reference) for acute kidney failure but did not reach significance for hyperkalaemia (OR: 18.51 [0.98-350.44]).

Figure 2:  Beta-blocker adjusted predictive interval plot (acute kidney failure). CAN, canrenone; PLA, Placebo; EPLE, eplerenone; SPIRO, spironolactone.




Figure 3: Cumulative ranking probability plots for beta-blocker adjusted SUCRA-values (acute kidney failure). CAN, canrenone; PLA, Placebo; EPLE, eplerenone; SPIRO, spironolactone.

Figure 4: Beta-blocker adjusted predictive interval plot (hyperkalaemia). CAN, canrenone; PLA, Placebo; EPLE, eplerenone; SPIRO, spironolactone.

Figure 5: Cumulative ranking probability plots for beta-blocker adjusted SUCRA-values (hyperkalaemia). CAN, canrenone; PLA, Placebo; EPLE, eplerenone; SPIRO, spironolactone.

Conclusion: The risk-benefit ratio of spironolactone with respect to acute renal failure and hyperkalaemia appears favourable as compared to eplerenone.