Applicability of mineralocorticoid receptor antagonist trials to real-world patients with heart failure across the ejection fraction spectrum
B N Beer, F Lindberg, L Benson, G SavareseAbstract
Aims
To evaluate the applicability of evidence from 8 past and ongoing mineralocorticoid receptor antagonist (MRA) trials to real-world patients with heart failure (HF) across the ejection fraction (EF) spectrum.
Methods
Patients enrolled in the Swedish Heart Failure Registry (SwedeHF) 2017-2023 were assessed for eligibility for RALES, EMPHASIS-HF, TOPCAT, FINEARTS-HF, REDEFINE-HF, FINALITY-HF, SPIRIT-HF, and SPIRRIT-HFpEF. Eligibility was defined as meeting all enrolment criteria for each trial. Patients with preserved (HFpEF) or mildly reduced EF (HFmrEF) were considered for TOPCAT, FINEARTS-HF, REDEFINE-HF, SPIRIT-HF, and SPIRRIT-HFpEF; those with reduced EF (HFrEF) for RALES, EMPHASIS-HF, and FINALITY-HF.
Results
58,915 patients with HF, 26.6% HFpEF, 28.7% HFmrEF, 44.7% HFrEF. 56.5% were eligible for >=1 trial (median age 75 years; eligible 77 vs ineligible 72 years). By HF phenotype, 56.5% with HFmrEF/HFpEF were eligible for ≥1 trial, and the same proportion (56.5%) applied to HFrEF. Eligibility varied widely between trials: 41.6% for TOPCAT (most restrictive criterion: prior HF hospitalisation [HHF] or NT-proBNP threshold), 31.9% FINALITY-HF (no prior steroidal MRA), 30.5% FINEARTS-HF (any diuretics), 26.8% SPIRRIT-HFpEF (loop diuretics), 19.9% EMPHASIS-HF (NYHA class II), 17.1% RALES (NYHA class III/IV), 9.0% REDEFINE-HF (hospitalised), and 5.8% SPIRIT-HF (prior HHF). It was predominantly limited by enrichment criteria reflecting HF severity – such as prior HHF, elevated NT-proBNP, or loop diuretics use – rather than by safety-related exclusions (severe renal dysfunction or hyperkalaemia). Compared with ineligible patients, those eligible had higher risks of HHF or cardiovascular (CV) death (rate ratio [RR] 1.38, 95%CI 1.35-1.42; p<0.001), but also of non-CV death (RR 1.10, 95%CI 1.04-1.16; p<0.001). In the subset of patients enrolled in 2022/23, trial eligibility declined: 40.3% for TOPCAT, 26.0% FINEARTS-HF, 25.2% FINALITY-HF, 21.4% SPIRRIT-HFpEF, 20.2% EMPHASIS-HF, 14.4% RALES, 5.1% REDEFINE-HF, and 4.3% SPIRIT-HF. This finding reflects, among other factors, lower median NYHA classes, median NT-proBNP levels, and less frequent use of diuretics in a more contemporary HF population. Patients in SwedeHF eligible for >=1 trial were substantially older (median 77 years) than those actually enrolled in any of the past trials, i.e. RALES (65 years), EMPHASIS-HF and TOPCAT (69 years), and FINEARTS-HF (72 years).
Conclusions
Most real-world patients irrespective of HF phenotype are eligible for >=1 MRA trial. Trial criteria selected for more severe HF and higher CV risk, but also for greater comorbidity and a slightly increased risk of non-CV events, underscoring the difficulty of separating CV from competing non-CV risks in trial design. In past trials, eligible real-world patients differed substantially from those actually enrolled, indicating limited generalisability beyond protocolised selection criteria.Table 1Figure 1