Key Points:
- Steroidal mineralocorticoid receptor antagonists (sMRA), such as spironolactone, are effective in heart failure with reduced ejection fraction (HFrEF), but their role in heart failure with preserved ejection fraction (HFpEF) and heart failure with mildly reduced ejection fraction (HFmrEF) remains uncertain.
- SPIRIT-HF randomized 730 patients with symptomatic HFpEF or HFmrEF across four European countries to spironolactone or placebo. The primary endpoint was the composite of cardiovascular (CV) death and total (first and recurrent) heart failure (HF) hospitalizations at 24 months.
- Spironolactone did not reduce the primary composite endpoint compared to placebo (Rate Ratio 1.18 [95% CI: 0.72–1.92]; p=0.512) and was associated with a significantly higher rate of total hospitalizations for any cause, as well as higher rates of hypotension, hyperkalemia, and renal events.
- However, SPIRIT-HF was significantly impacted by the COVID-19 pandemic, enrolling approximately 50% fewer patients than planned and experiencing high rates of study drug discontinuation. As such, the trial was underpowered to adequately test its primary hypothesis. Future trials are needed to further define the role of MRA therapy in HFpEF and HFmrEF.
Heart failure with preserved ejection fraction (HFpEF) and heart failure with mildly reduced ejection fraction (HFmrEF) contribute substantially to the global burden of heart failure, yet evidenced-based therapies for these conditions remain limited. Steroidal mineralocorticoid receptor antagonists (sMRA), including spironolactone, have demonstrated clear clinical benefit in HFrEF, but their efficacy in HFpEF and HFmrEF remains uncertain. The largest prior trial of sMRA in this population, TOPCAT, yielded mixed results overall, though a sub-analysis suggested benefit among participants enrolled in the Americas.
On March 29th, 2026, the results of “SPIRonolactone In the Treatment of Heart Failure (SPIRIT-HF)” were presented at the American College of Cardiology Scientific Sessions in New Orleans, LA. The purpose of this trial was to determine whether spironolactone improves cardiovascular outcomes in patients with stable HFmrEF or HFpEF at high risk for cardiovascular events.
SPIRIT-HF was a prospective, publicly funded, multicenter, double-blind, placebo-controlled outcome trial conducted at 56 sites across four European countries (Germany 81%, the Netherlands 7%, Austria 7%, and France 5%) [NCT04727073] . Patients aged ≥50 years with symptomatic heart failure (NYHA class II–IV), LVEF ≥40% (stratified by 40–49% and ≥50%), echocardiographic evidence of structural and/or functional cardiac abnormalities, and either a HF hospitalization within the past 12 months or elevated NT-proBNP levels were randomized in a 1:1 ratio to spironolactone or placebo. The initial target dose of spironolactone was 50 mg daily, which was amended during the trial to 25 mg daily, with dose adjustment based on renal function (25 mg every other day for eGFR <50 mL/min/1.73m²). The primary endpoint was the composite of CV death and total (first and recurrent) HF hospitalizations at 24 months, analyzed using a semiparametric model for recurrent events.
Overall, 893 patients were screened, 743 enrolled, and 730 were randomized, 366 to spironolactone and 364 to placebo. This was significantly lower than the sample size of 1564 total patients based on their power analysis. The mean age was 76.3 years, 52% were female, and 96% were White. Approximately 80% of participants had HFpEF (LVEF ≥50%) and 20% had HFmrEF (LVEF 40–49%); the median NT-proBNP was 970 pg/mL and 65% were NYHA class II. Background medical therapy included beta-blockers (76%), diuretics (83%), and ACE inhibitors or ARBs (70%). Nearly 70% had atrial fibrillation, which may explain the high beta blocker use. There was high loss to follow-up due to the COVID-19 pandemic, with only 374 total patients reporting at 24 months. There was also a high rate of study drug discontinuation in both groups, with 50% of spironolactone patients having stopped or taken less than 20% of the study drug by month 21, compared to month 31 for placebo.
Within these limitations, spironolactone did not reduce the primary composite endpoint of CV death and total HF hospitalizations at 24 months compared to placebo (RR 1.18 [95% CI: 0.72–1.92]; p=0.51). Sensitivity analyses restricted to the first 12 months (RR 0.93 [95% CI: 0.49–1.78]) and on-treatment analyses (RR 0.94 [95% CI: 0.49–1.79]) were also non-significant. No heterogeneous treatment effects were observed across prespecified subgroups. An analysis of time-varying rate ratios revealed a significant time trend (p=0.025): the rate ratio favored spironolactone early in the trial but crossed above 1.0 from approximately month 5 onward, coinciding with rising rates of study drug discontinuation.
Among secondary endpoints, spironolactone was associated with a significantly higher rate of total hospitalizations for any cause compared to placebo (RR 1.40 [95% CI: 1.07–1.83]; p=0.013), and a trend toward higher cardiovascular hospitalizations (RR 1.36 [95% CI: 0.99–1.86]; p=0.055). There was no significant difference in HF hospitalizations alone, CV death, or all-cause death between the two groups. Regarding safety, spironolactone was associated with significantly higher rates of hypotension (RR 2.26 [95% CI: 1.22–4.17]; p=0.009), hyperkalemia (RR 3.49 [95% CI: 1.49–8.19]; p=0.004), and renal events (RR 1.75 [95% CI: 1.07–2.85]; p=0.025) compared to placebo.
A prespecified meta-analysis pooling SPIRIT-HF with the TOPCAT-Americas cohort found a non-significant combined rate ratio of 0.84 (95% CI: 0.70–1.01) for CV death or HF hospitalization over the complete follow-up period, and 0.76 (95% CI: 0.58–0.98) when restricted to the first 12 months. Bayesian analyses incorporating between-trial heterogeneity yielded wide credible intervals, reflecting substantial uncertainty.
Overall, the trial enrolled approximately 50% fewer patients than the planned sample size of 1,564, captured fewer than 75% of the assumed number of events, and experienced high rates of study drug discontinuation in both arms during the pandemic. The dose amendment from 50 mg to 25 mg during the trial may have also contributed to reduced efficacy. The investigators acknowledged that SPIRIT-HF was underpowered to adequately test its primary hypothesis. The population was predominantly European and White, which may also limit generalizability. Additionally, only 17% of patients were on SGLT2 inhibitors and 5% on an ARNI, which have since become standard therapy for HFpEF.
Frank Edelmann, MD, Professor for Cardiovascular Prevention at the Deutsches Herzzentrum der Charité in Berlin, Germany, concluded: “We need to be careful about the interpretation of these findings because ultimately the trial was a little too small, but there are some issues regarding safety and efficacy, and it is very important for the community to have this data and to discuss it. On efficacy, the study was negative or perhaps slightly inconclusive, but the clear increase in total hospitalizations and adverse events warrants more attention.”