Summary: In 108 patients with heart failure with mildly reduced or preserved ejection fraction (HFmrEF/HFpEF), the open-label, blinded-endpoint crossover SOGALDI-PEF trial found that adding spironolactone to dapagliflozin lowered LogNT-proBNP by a modest 11% relative to dapagliflozin alone (-0.11 Log-units; 95% CI -0.22 to -0.01; P=0.035). The effect was borderline statistically significant and was offset by a greater decline in eGFR and higher serum potassium. NT-proBNP is a surrogate biomarker, not a clinical outcome.
PICO Summary
| Element | Detail |
|---|---|
| Population | 108 patients with HFmrEF or HFpEF; median age 76 years, 57% women, 45% with diabetes; multicentre, Portugal (SOGALDI-PEF, NCT05676684). |
| Intervention | Dapagliflozin plus spironolactone for 12 weeks (crossover design; all 108 patients received both sequences). |
| Comparison | Dapagliflozin alone for 12 weeks (within-patient comparator; all 108 patients). |
| Outcome | Primary: LogNT-proBNP fell by -0.11 Log-units (95% CI -0.22 to -0.01; P=0.035), an ~11% relative reduction. Secondary: odds of >=20% NT-proBNP reduction OR 2.27 (95% CI 1.16-4.44; P=0.016); systolic BP -5.2 mmHg (95% CI -8.4 to -2.0); Log urinary albumin-to-creatinine ratio -0.32 (95% CI -0.54 to -0.11). Safety trade-offs: eGFR -6.4 mL/min/1.73m2 (95% CI -8.3 to -4.4); serum potassium +0.32 mmol/L (95% CI 0.23-0.41); potassium >5.5 mmol/L in 5 (4.8%) vs 1 (0.9%). No hard clinical-outcome data (hospitalisation, mortality). |
SOGALDI-PEF
Crossover RCT · HFmrEF/HFpEF · 12 weeks
Adding spironolactone to dapagliflozin cut NT-proBNP by ~11% (borderline), at the cost of lower eGFR and higher potassium. A surrogate signal, not proven clinical benefit.
Expert Commentary
SOGALDI-PEF is a small, hypothesis-generating crossover trial, and its result should be read as a biomarker signal rather than proof of clinical benefit. Adding spironolactone to dapagliflozin lowered NT-proBNP by roughly 11%, but the effect was borderline, with the upper confidence bound (-0.01 Log-units) sitting almost on the null and a P value of 0.035. The accompanying reductions in blood pressure and albuminuria are biologically coherent with mineralocorticoid receptor blockade, yet they were bought at the cost of a clinically meaningful eGFR fall and a rise in serum potassium, with more patients crossing the 5.5 mmol/L threshold. A central caveat is the open-label, blinded-endpoint (PROBE) design: although the laboratory endpoints were adjudicated blind, treatment was unblinded, which can bias adherence and co-intervention. The trial was also powered only for a surrogate, was conducted at sites in a single country, and reported no hospitalisation or mortality data. Can I use this with my patients? Not yet as a routine combination; the data do not establish that adding spironolactone improves outcomes, and the renal-potassium trade-off matters in the older, comorbid HFpEF population studied here. It may inform shared decisions in a selected patient with preserved renal function and low-normal potassium who tolerates monitoring. A larger, outcome-powered trial is needed before this combination is recommended in practice.
References
Ferreira JP, Vasques-Nóvoa F, Saraiva F, et al. Sodium-Glucose Cotransporter 2 Inhibitor With and Without an Aldosterone Antagonist for Heart Failure With Preserved Ejection Fraction: The SOGALDI-PEF Trial. J Am Coll Cardiol. 2025;86(5):320-333. doi:10.1016/j.jacc.2025.05.033