Advances in the Pharmacological Treatment of Heart Failure With Preserved Ejection Fraction

Abstract

Heart failure with preserved ejection fraction (HFpEF) represents approximately half of all heart failure cases and poses a growing global health challenge driven by an ageing population and an increasing comorbidity burden. Once regarded as a condition without effective, evidence-based therapy, HFpEF has undergone a paradigm shift in recent years. Advances in the understanding of its complex pathophysiology have highlighted the multifactorial interplay between systemic inflammation, endothelial dysfunction, and metabolic derangements. The introduction of sodium-glucose cotransporter 2 inhibitors has transformed the HFpEF therapeutic landscape, following large-scale trials such as EMPEROR-Preserved and DELIVER demonstrating consistent reductions in mortality/morbidity in this patient population. More recently, the non-steroidal mineralocorticoid receptor antagonist finerenone, tested in the FINEARTS-HF trial, was also shown to improve mortality/morbidity in HFpEF, marking a further milestone in disease-modifying therapy. Further, glucagon-like peptide-1 receptor agonists (GLP-1 RAs) and a dual gastric inhibitory polypeptide analogue/GLP-1 RA (tirzepatide) have shown to reduce body weight and improve quality of life and mortality/morbidity in the obese-HFpEF phenotype, suggesting the need of additional tailoring of HFpEF therapy based on specific patient profiles. Despite these advances, HFpEF remains a frequently underdiagnosed syndrome, with diagnostic uncertainty often delaying therapy. Comprehensive management of comorbidities and systematic implementation of guideline-directed medical therapy remain crucial to improve patient outcomes. This narrative review provides an updated overview of the pathophysiological mechanisms, diagnostic approaches, and evolving pharmacological strategies shaping the modern management of HFpEF.

Graphical Abstract

HEART FAILURE WITH PRESERVED EJECTION FRACTION (HFpEF) IN 2025: GLOBAL BURDEN AND EVOLVING THERAPEUTIC LANDSCAPE

Heart failure (HF) is a global pandemic, characterised by high morbidity and mortality, impaired quality of life, and substantial healthcare costs.¹⁾ HF is classified into distinct subgroups based on left ventricular ejection fraction (EF) at the time of diagnosis. An EF ≤40% defines heart failure with reduced ejection fraction (HFrEF), 41–49% heart failure with mildly reduced ejection fraction (HFmrEF), and ≥50% HFpEF.²⁾ Heart failure with improved ejection fraction (HFimpEF) refers to patients with a baseline EF ≤40%, a ≥10-point increase from baseline, and a subsequent EF >40%.²⁾ About half of HF patients are diagnosed with HFpEF, corresponding to an estimated 32 million individuals worldwide.³,⁴,⁵⁾ Nevertheless, the true prevalence is likely higher, as 35–75% of patients fulfilling diagnostic criteria for HFpEF are estimated remaining unrecognised in clinical practice.⁶,⁷⁾ The prevalence of HFpEF has steadily increased over the past decade, and is expected to continue rising in the near future, driven by ageing populations, increased awareness, and the growing burden of comorbidities such as hypertension, obesity, atrial fibrillation, diabetes, and chronic kidney disease (CKD).¹,⁸,⁹,¹⁰⁾ The incidence of HFpEF ranges from 1 to 4 cases per 1,000 person-years, depending on cohort characteristics and study period. It continues to increase even in regions where overall HF incidence has stabilised or declined, likely reflecting improved management of other cardiovascular conditions, e.g. the treatment of ischaemic heart disease.¹¹,¹²,¹³⁾ Nevertheless, prognosis in HFpEF remains poor. In-hospital mortality ranges from 2% to 5%, annual mortality is approximately 15% and increases with age, whereas 5-year rates of rehospitalisation and death reach up to 80% and 50%, respectively.³⁾

Despite these worrisome trends, HFpEF was long considered a clinical entity with few effective evidence-based treatments, largely due to its phenotypic heterogeneity impacting the likelihood of success of randomised controlled trials, and the incomplete understanding of the pathophysiological mechanisms.¹⁴,¹⁵⁾ In recent years, however, emerging evidence from prospective, randomised trials has substantially expanded the range of pharmacological options, providing new opportunities to modify disease progression in selected patients with HFpEF.¹⁶⁾ Sodium-glucose cotransporter 2 inhibitors (SGLT2i) have been approved to reduce cardiovascular mortality or HF hospitalisations in patients with HF across the EF spectrum.¹⁷,¹⁸,¹⁹,²⁰,²¹⁾ The FINEARTS-HF trial has provided the first evidence that the non-steroidal mineralocorticoid receptor antagonist (MRA) finerenone impacts hard outcomes (cardiovascular death or total HF events) in patients with HFmrEF and HFpEF, further expanding treatment opportunities in this population.²²⁾ In parallel, there has been growing interest in targeting comorbidities, systemic inflammation, and mechanisms of myocardial remodelling.²³,²⁴⁾

Accordingly, this narrative review aims to provide a contemporary overview of HFpEF, summarising key aspects of its pathophysiology and diagnosis, and outlining recent pharmacological advances, their clinical implications, and directions for future research.

PATHOPHYSIOLOGY

HFpEF was initially considered a disorder mainly characterised by diastolic dysfunction secondary to hypertension, leading to left ventricular hypertrophy with a relatively small and stiff ventricle, ultimately progressing to overt HF.²⁵,²⁶⁾ Over time, however, the understanding of the clinical phenotype of HFpEF has evolved from being considered an isolated hypertensive heart disease in older adults to a more complex condition contributing to define the cardio-renal metabolic disease, as characterised by the interplay of multiple comorbidities.²⁷,²⁸⁾ This shift has broadened the understanding of HFpEF to a multifactorial syndrome involving systemic inflammation, endothelial dysfunction, altered myocardial energetics, and abnormalities in skeletal muscle (Figure 1).²⁹⁾

Figure 1. Pathophysiological mechanisms in HFpEF and therapeutic targets.

HFpEF = heart failure with preserved ejection fraction; RAAS = renin-angiotensin-aldosterone system; SGLT2 = sodium-glucose cotransporter 2; MR = mineralocorticoid receptor; GLP-1 = glucagon-like peptide-1.

A key pathological process in HFpEF is passive myocardial stiffness, defined as reduced ventricular compliance during diastole.³⁰⁾ In the healthy heart, passive stiffness facilitates efficient ventricular filling and contributes to contractile force generation through the Frank-Starling mechanism.³⁰⁾ In HFpEF, however, excessive stiffness impairs diastolic filling and limits the ability to augment cardiac output during exertion.³⁰⁾ The myocardial basis of these abnormalities remains incompletely understood, but clinical studies suggest that both intracellular and extracellular mechanisms, such as titin and extracellular matrix remodelling, may contribute to this process.³⁰⁾

Beyond increased passive myocardial stiffness, patients with HFpEF also exhibit impaired active relaxation, particularly during exertion.³¹⁾ Under physiological conditions, active relaxation is an energy-dependent process that relies on the reuptake of cytosolic calcium via sarcoplasmic/endoplasmic reticulum calcium ATPase pump during diastole.³¹⁾ In HFpEF, this process is delayed due to impaired calcium handling and reduced energetic reserve, leading to slower ventricular relaxation, elevated filling pressures, and diastolic dysfunction.³¹⁾

In 2013, Paulus and Tschöpe³²⁾ proposed a new paradigm in which comorbidities such as obesity, diabetes, hypertension, chronic obstructive pulmonary disease, anaemia, and CKD promote a systemic proinflammatory state that induces coronary microvascular endothelial dysfunction. This endothelial impairment reduces nitric oxide bioavailability and downstream cyclic guanosine monophosphate (cGMP) and protein kinase G (PKG) signalling within cardiomyocytes, ultimately promoting hypertrophy, increased stiffness, and interstitial fibrosis.³²⁾ However, therapies targeting this pathway have not yet demonstrated efficacy in clinical trials, although it remains unclear whether these interventions effectively restore intracellular cGMP-PKG signaling.³³,³⁴⁾ More recent studies have expanded this paradigm into the concept of metabolic inflammation (“metaflammation”), in which obesity and nutrient overload trigger chronic microvascular inflammation, oxidative stress, and altered myocardial energy metabolism.³⁵⁾ The primary source of this inflammation in HFpEF remains uncertain. Adipocytes and adipose tissue macrophages may release proinflammatory adipokines such as interleukin-6, while increased visceral fat has been independently linked to higher HFpEF risk and disease severity.³⁶⁾ Emerging evidence also suggest that SGLT2i may counteract these processes by enhancing nutrient deprivation signalling, thereby improving cellular quality control and reducing inflammation and oxidative stress.³⁷⁾

Increasing attention has also been directed towards overactivation of the renin-angiotensin-aldosterone system (RAAS) in HFpEF.³⁸⁾ Aldosterone, the final effector hormone of this cascade, binds to the mineralocorticoid receptor (MR), a cytoplasmic steroid receptor expressed in the kidney, heart, and vasculature.³⁹,⁴⁰⁾ Under physiological conditions, MR activation maintains fluid, electrolyte, and haemodynamic homeostasis.³⁹⁾ However, chronic overstimulation leads to sodium and water retention, and activates proinflammatory and profibrotic pathways, contributing to myocardial and endothelial injury.³⁹⁾ Although MR activation is a well-recognised driver of inflammation, fibrosis, and vascular dysfunction in HFrEF, its specific contribution in HFpEF remains less established yet mechanistically plausible.³⁹,⁴¹⁾ Experimental data suggest that MRAs may mitigate diastolic dysfunction and improve outcomes in HFpEF, possibly counteracting RAAS-mediated cardiac fibrosis and hypertrophy.³⁸,⁴²⁾ This biological rationale supports further exploration of MR blockade as a therapeutic target in HFpEF.

In recent years, the “multiple phenotype hypothesis” has been proposed to explain the heterogeneity of HFpEF, suggesting that patients may reach a common haemodynamic profile of elevated left ventricular filling pressures through distinct underlying mechanisms.²⁹⁾ Among the proposed pathophysiological phenotypes, obesity has emerged as one of the most prevalent, affecting 30–40% of patients with HFpEF.⁴³⁾ Obesity-driven HFpEF results from the interplay of haemodynamic alterations, systemic inflammation, and neurohormonal dysregulation.⁴⁴⁾ Volume expansion, increased afterload, and excess epicardial fat create a haemodynamic milieu that elevates cardiac filling pressures and promotes concentric remodelling and diastolic stiffness.⁴⁵⁾ In parallel, chronic low-grade inflammation driven by dysfunctional adipose tissue transforms fat from a passive energy store into a metabolically active organ that releases proinflammatory cytokines, disrupts endothelial function, and promotes myocardial fibrosis.⁴⁶⁾ Obesity-related HFpEF is also characterised by neurohormonal activation, as persistent sympathetic drive and RAAS stimulation increase vascular resistance, sodium retention, and myocardial hypertrophy.⁴⁴⁾ Glucagon-like peptide-1 receptor agonists (GLP-1 RAs), incretin-based therapies originally developed for the treatment of type 2 diabetes, have recently emerged as a promising class of agents for patients with HFpEF and obesity.⁴⁴⁾ Beyond glycaemic control and weight loss, these agents exert pleiotropic actions on cardiovascular, metabolic, and inflammatory pathways, including alleviating systemic inflammation, reducing epicardial fat, and lowering preload and afterload, thereby improving cardiac structure and function.⁴⁴⁾ Further studies are warranted to clarify whether these benefits translate into improved outcomes in HFpEF.

DIAGNOSIS

Diagnosing HFpEF can be challenging and relies upon the presence of symptoms and/or signs of HF, an EF ≥50%, and objective evidence of diastolic dysfunction or elevated filling pressures, including elevated natriuretic peptide levels.²⁾ Common symptoms such as dyspnoea and fatigue are often non-specific and may overlap with those of common comorbidities that limit exercise tolerance, including metabolic disorders, valvular disease, pulmonary disease, and anaemia.⁴⁷,⁴⁸⁾ Signs of congestion may be evident on clinical examination, but are frequently absent in ambulatory patients.⁴⁹⁾ Natriuretic peptides may support the diagnosis, but up to one-third of patients with HFpEF have natriuretic peptide values below conventional diagnostic thresholds, particularly in the presence of obesity or reduced left ventricular wall stress.⁴⁵,⁵⁰⁾ Echocardiography remains central to HFpEF diagnosis and can reveal markers of congestion, an increased E/e′ ratio, left atrial enlargement, reduced atrial strain, elevated estimated pulmonary artery pressures, or inferior vena cava distension.⁵¹⁾ However, the absence of these findings does not exclude HFpEF, as filling pressures may be normal at rest but increase during exercise.⁵²⁾ When diagnostic uncertainty persists, exercise stress testing or invasive haemodynamic assessment during right heart catheterisation may be required.⁵³⁾ Although the latter can provide a definitive diagnosis, it remains technically demanding, costly, and not universally available in clinical practice.⁵³⁾ To improve diagnostic precision, 2 score-based algorithms (H₂FPEF and HFA-PEFF) have been proposed, integrating clinical, imaging, and biomarker data to estimate the likelihood of HFpEF in patients with unexplained dyspnoea.⁴⁷,⁵⁴⁾ Both scores have been validated as accurate tools to discriminate HFpEF from non-cardiac causes of dyspnoea, but they differ in performance. The H₂FPEF score, which requires fewer input variables, has shown greater sensitivity, overall accuracy, and discriminative ability than the HFA-PEFF score in large multicentre studies with invasive haemodynamics as reference.⁵⁵⁾ However, it is noteworthy that the H₂FPEF score was originally derived and validated in patients with unexplained dyspnoea. In patients with overt signs and symptoms of HF and elevated natriuretic peptides, a low H₂FPEF score does not exclude HFpEF, but may instead indicate mimics of HFpEF such as infiltrative cardiomyopathies.²⁹⁾ Overall, HFpEF remains a complex and frequently under-recognised syndrome, with diagnostic uncertainty often delaying treatment and underscoring the need for more sensitive, standardised, and widely accessible approaches to diagnosis.

ESTABLISHED PHARMACOLOGICAL LANDSCAPE

SGLT2i

For decades, the management of HFpEF has focused on managing comorbidities and symptoms, mainly with diuretics.¹⁴,⁵³⁾ Only recently, the therapeutic landscape has evolved with the introduction of SGLT2i, which have demonstrated consistent benefits in patients with HF across the EF spectrum, including those with worsening or acute HF.²¹,⁵⁶,⁵⁷,⁵⁸⁾ Initially developed for the treatment of type 2 diabetes, these agents were later found to improve hard cardiovascular outcomes in HFrEF, including a reduction in HF hospitalisations and, unexpectedly, cardiovascular death.¹⁷,¹⁸⁾ In parallel, SGLT2i have also been shown to slow kidney function decline and improve renal and HF outcomes in patients with CKD. In the SCORED trial, which enrolled patients with type 2 diabetes, CKD across the full range of albuminuria, and additional cardiovascular risk factors, sotagliflozin significantly lowered the risk of deaths from cardiovascular causes, HF hospitalisations and urgent visits for HF.⁵⁹⁾ In the DAPA-CKD and EMPA-KIDNEY trials, which included CKD patients irrespective of diabetes status, dapagliflozin and empagliflozin consistently reduced kidney disease progression, end-stage kidney disease, death from renal causes, and cardiovascular death, supporting their broad efficacy across CKD populations.⁶⁰,⁶¹⁾

Subsequent analyses, focusing on patients with HFpEF, further confirmed that the benefits of SGLT2 inhibition extend beyond glycaemic control in type 2 diabetes (Figure 1).⁶²,⁶³⁾ In SOLOIST-WHF, which enrolled patients with worsening HF across the full range of EF, SGLT2i significantly reduced the total number of deaths from cardiovascular causes and hospitalisations and urgent visits for HF.⁵⁷⁾ In EMPULSE, which included patients hospitalised for acute HF regardless of EF, SGLT2 inhibition increased clinical benefit, defined as a hierarchical composite of all-cause death, number of HF events, time to first HF event, or a ≥5-point improvement in the Kansas City Cardiomyopathy Questionnaire (KCCQ) total symptom score at 90 days, without any heterogeneity of the treatment effect according to EF.⁵⁸⁾

In the EMPEROR-Preserved trial, 5,988 patients with symptomatic HF and EF >40%, were randomised to receive empagliflozin or placebo on top of standard of care.²¹⁾ The primary endpoint, a composite of time to cardiovascular death or first HF hospitalisation, occurred in 13.8% of patients in the empagliflozin group and in 17.1% of those in the placebo group, over a median of 26.2 months, accounting for a statistically significant relative risk reduction of 21% (hazard ratio [HR], 0.79; 95% confidence interval [CI], 0.69–0.90).²¹⁾ The benefit was mainly driven by a 29% lower risk of HF hospitalisations (HR, 0.71; 95% CI, 0.60–0.83), whereas cardiovascular mortality was not statistically significantly reduced (HR, 0.91; 95% CI, 0.76–1.09).²¹⁾ Empagliflozin also decreased the total number of HF hospitalisations (HR, 0.73; 95% CI, 0.61–0.88), slowed the decline in estimated glomerular filtration rate (eGFR, −1.25 vs. −2.62 mL/min/1.73 m² per year), and improved quality of life as measured by the KCCQ.²¹⁾ Benefits were consistent across prespecified subgroups, including diabetes, sex, renal function, blood pressure, age, and EF category (<50% and 50–59%), with only modest attenuation at EF ≥60%.²¹⁾ In post hoc analyses, empagliflozin also reduced serum uric acid and the need for diuretic initiation or dose escalation.⁶⁴,⁶⁵⁾

In a pooled EMPEROR-Reduced and EMPEROR-Preserved analysis, empagliflozin further reduced the incidence of hyperkalaemia and the need for potassium binders compared with placebo (6.5% vs. 7.7%; HR, 0.82; 95% CI, 0.71–0.95), without increasing the risk of hypokalaemia, suggesting an additional mechanism through which SGLT2 inhibition may facilitate the implementation and continuation of guideline-directed medical therapy.⁶⁶⁾ Moreover, in a further analysis of EMPEROR-Reduced and EMPEROR-Preserved, empagliflozin reduced the risk of major adverse renal outcomes (i.e., profound and sustained decreases in eGFR or renal replacement therapy) by 49% in EMPEROR-Reduced (HR, 0.51; 95% CI, 0.33–0.79).⁶⁷⁾ However, no significant benefit was observed in EMPEROR-Preserved (HR, 0.95; 95% CI, 0.73–1.24), despite a consistent slowing of eGFR decline in both trials, suggesting that EF may influence the effects of empagliflozin on major renal outcomes.⁶⁷⁾

The cardiovascular benefits of SGLT2i were subsequently confirmed and extended in the DELIVER trial, which enrolled 6,263 patients with symptomatic HF and an EF >40%.¹⁹⁾ Participants were randomised to receive dapagliflozin or placebo on top of standard of care and were followed for a median of 2.3 years. Dapagliflozin significantly reduced the primary composite endpoint of cardiovascular death or worsening HF, defined as HF hospitalisation or an urgent out-patient visit requiring intravenous therapy (16.4% vs. 19.5%; HR, 0.82; 95% CI, 0.73–0.92), driven mainly by a reduction in HF hospitalisations (HR, 0.79; 95% CI, 0.69–0.91), whereas cardiovascular mortality did not differ significantly between groups (HR, 0.88; 95% CI, 0.74–1.05). Total (first and recurrent) events of cardiovascular death or worsening HF were also fewer with dapagliflozin (rate ratio, 0.77; 95% CI, 0.67–0.89). Dapagliflozin also improved quality of life, with a significant increase in KCCQ total symptom score versus placebo (win ratio, 1.11; 95% CI, 1.03–1.21; mean difference +2.4 points, 95% CI, 1.5–3.4). Consistent with EMPEROR-Preserved, the efficacy of dapagliflozin for the primary outcome did not vary across all prespecified subgroups, including diabetes status, sex, baseline renal function, age, blood pressure, and was maintained across the entire EF spectrum, including patients with EF ≥60%.¹⁹⁾ In a prespecified renal analysis, baseline kidney function did not modify the benefit of dapagliflozin (eGFR ≥60 mL/min/1.73 m²: HR, 0.84; 95% CI, 0.70–1.00; eGFR 45–59 mL/min/1.73 m²: HR, 0.68; 95% CI, 0.54–0.87; eGFR <45 mL/min/1.73 m²: HR, 0.93; 95% CI, 0.76–1.14; p-interaction=0.16).⁶⁸⁾ Dapagliflozin also attenuated the rate of eGFR decline compared with placebo (difference +0.5 mL/min/1.73 m² per year; 95% CI, 0.1–0.9), without increasing the incidence of serious renal adverse events.⁶⁸⁾ In post hoc analyses, dapagliflozin reduced new loop diuretic initiations, dose escalations, and mean diuretic requirements over time, and was associated with early and sustained improvements in health-related quality of life, particularly among patients with greater frailty.⁶⁹,⁷⁰⁾

Compared with EMPEROR-Preserved, DELIVER enrolled a broader and more heterogeneous HFpEF population, including patients who were recently hospitalised or had improved EF, defined as a prior EF ≤40% that had recovered to >40% at enrollment.¹⁹,²¹⁾ DELIVER also adopted more inclusive definition of worsening HF that incorporated urgent out-patient visits requiring intravenous therapy, thereby more closely reflecting real-world patterns of decompensation. Despite these methodological differences, both trials demonstrated remarkably similar relative risk reductions (~20%) in the primary composite endpoint, with consistent benefits across EF strata, clinical subgroups, and comorbidity burden. When pooled (n≈12,000), results from EMPEROR-Preserved and DELIVER confirmed a ~20% relative risk reduction in cardiovascular death or HF hospitalisation with SGLT2i.⁷¹⁾ Accordingly, SGLT2i are now recommended for HFpEF as class IIa B-R in the 2022 American Heart Association (AHA), American College of Cardiology (ACC), and Heart Failure Society of America (HFSA) guidelines, which were based primarily on the results of EMPEROR-Preserved, and class I A in the 2023 European Society of Cardiology (ESC) Focused Update of the Guidelines, which incorporated consistent evidence from both EMPEROR-Preserved and DELIVER.²⁰,⁷²⁾

Angiotensin receptor-neprilysin inhibitors (ARNi)

Evidence for ARNi in HFpEF has been less definitive. In the PARAGON-HF trial, which enrolled 4,822 patients with symptomatic HF and an EF ≥45%, sacubitril/valsartan did not significantly reduce the primary composite endpoint of total HF hospitalisations and cardiovascular death compared with valsartan alone (rate ratio, 0.87; 95% CI, 0.75–1.01). Although the trial was neutral for the primary endpoint, there was a suggestion of heterogeneity of the treatment effect with possible benefit among patients with lower EF (<57%) and women.⁷³⁾ Sacubitril/valsartan also led to modest but statistically significant improvements in health-related quality of life, reflected by a smaller decline in KCCQ score compared with valsartan (1.6 vs. 2.6 points; between-group difference 1.0 point, 95% CI, 0.0–2.1) and a higher proportion of patients achieving a ≥5-point improvement (33% vs. 29.6%; odds ratio, 1.30; 95% CI, 1.04–1.61). In a prespecified analysis, the composite renal outcome (i.e., time to first occurrence of either ≥50% eGFR decline, end-stage renal disease, or death from renal causes) occurred in 1.4% of patients receiving sacubitril/valsartan and 2.7% receiving valsartan, accounting for a statistically significant 50% reduction in risk (HR, 0.50; 95% CI, 0.33–0.77).⁷⁴⁾ The treatment effect was consistent across baseline eGFR strata, and the annual eGFR decline was slower with sacubitril/valsartan than with valsartan (−2.0 vs. −2.7 mL/min/1.73 m² per year), indicating a renoprotective effect of ARNi in HFpEF. In post hoc analyses, sacubitril/valsartan showed a trend towards improved outcomes in patients who initiated treatment in proximity to the index hospitalisation or had an EF ≤57% and eGFR ≤45 mL/min/1.73 m², although these findings remain exploratory and should be interpreted with caution.⁷⁵,⁷⁶⁾ Pooled analyses of PARADIGM-HF and PARAGON-HF, which included patients across the full EF spectrum, further support that the therapeutic effects of sacubitril/valsartan on HF hospitalisation vary by EF, with the greatest benefits observed in patients with EF below the normal range (approximately ≤55%).⁷⁷⁾ Based on these findings, the U.S. Food and Drug Administration expanded the indication for sacubitril/valsartan to all adults with chronic HF, emphasising that the strongest benefits occur in those with mildly reduced or low-normal EF.⁷⁸⁾ Consistently, the 2022 AHA/ACC/HFSA guideline indicates that ARNi may be considered to reduce HF hospitalisations in HFpEF, particularly among patients with EF values closer to the lower limit of the preserved range (class IIb B-R), whereas the latest ESC guidelines do not recommend its use in HFpEF, given the absence of conclusive evidence of benefit.²⁰⁾

Given these divergent recommendations, uncertainty remains regarding the optimal use of sacubitril/valsartan in patients with EF >40%. To address this gap, the PARAGLIDE-HF trial was conducted to further evaluate the efficacy and safety of sacubitril/valsartan as compared with valsartan in patients with HF and EF >40% following a recent worsening HF event.⁷⁹⁾ In 466 patients followed for a median of 5.9 months, sacubitril/valsartan achieved a greater reduction in N-terminal prohormone of brain natriuretic peptide (NT-proBNP) from baseline through weeks 4 and 8 (ratio of change, 0.85; 95% CI, 0.73–0.999). The reduction in NT-proBNP was evident as early as week 1 and was consistent across most subgroups. Nevertheless, the trial was not adequately powered to detect differences in hard outcomes, and sacubitril/valsartan did not significantly reduce the hierarchical secondary composite outcome of cardiovascular death, HF hospitalisation, urgent HF visits, and change in NT-proBNP (win ratio, 1.19; 95% CI, 0.93–1.52). However, sacubitril/valsartan was associated with a lower risk of worsening renal function (odds ratio, 0.61; 95% CI, 0.40–0.93). Consistent with prior evidence, treatment effects appeared more pronounced among patients with EF ≤60%, both for NT-proBNP reduction (0.78; 95% CI, 0.61–0.98) and for the hierarchical composite outcome (win ratio, 1.46; 95% CI, 1.09–1.95).

When participant-level data from PARAGLIDE-HF and PARAGON-HF were pooled, sacubitril/valsartan significantly reduced total worsening HF events and cardiovascular death in both the primary analysis of participants with recent worsening HF (rate ratio, 0.78; 95% CI, 0.61–0.99) and in the overall population (rate ratio, 0.86; 95% CI, 0.75–0.98).⁸⁰⁾ Benefits were greater in patients with EF ≤60%, and the risk of renal events was also lower (HR, 0.60; 95% CI, 0.44–0.83). Taken together, these findings provide more robust evidence to guide clinical decision-making on sacubitril/valsartan use in HFmrEF and HFpEF, but further trials are needed to better define which patients may derive the greatest benefit.

Beta-blockers, angiotensin-converting enzyme inhibitors (ACEi), and angiotensin II receptor blockers (ARBs)

Other neurohormonal agents have shown limited or inconsistent efficacy in HFpEF. Beta-blockers are not recommended specifically for the treatment of HFpEF, as supporting evidence is weak and emerging data indicate a potential increase in HF hospitalisations, especially in patients with EF >60%.⁵³,⁷²,⁸¹,⁸²⁾ Their use should therefore be limited to the management of conditions that coexist with HFpEF, such as ischaemic heart disease, hypertension, atrial fibrillation, or hyperthyroidism.⁵³,⁷²⁾

Similarly, renin-angiotensin system inhibitors, including ACEi and ARBs, have shown neutral or modest effects in HFpEF.⁸³⁾ In the CHARM-Preserved trial, which enrolled 3,023 patients with symptomatic HF and EF >40%, candesartan reduced the risk of HF hospitalisation compared with placebo (230 vs. 279 events; p=0.017) over a median follow-up of 36.6 months.⁸⁴⁾ However, these findings were not validated in the I-PRESERVE trial, which showed no impact of irbesartan on the primary composite endpoint of cardiovascular death or cardiovascular hospitalisation in 4,128 patients with symptomatic HF and EF ≥45% over a mean follow-up of 49.5 months.⁸⁵⁾ Accordingly, renin-angiotensin system inhibitors have no specific indication for HFpEF in the latest ESC guidelines, whereas the 2022 AHA/ACC/HFSA guidelines support consideration of ARBs (class IIb B-R) to reduce hospitalisations, particularly among patients with EF closer to the lower limit of the preserved range.²⁰,⁷²⁾

MRAs

The steroidal MRAs spironolactone and eplerenone constitute a key pharmacological therapy in HFrEF, based on landmark trials such as RALES and EMPHASIS-HF.⁸⁶,⁸⁷⁾ They have received respective class I A treatment recommendations by the ESC and the AHA/ACC/HFSA guidelines on the treatment of HF.⁵³,⁷²⁾

The role of steroidal MRAs in HFpEF remains uncertain and subject to ongoing debate. The TOPCAT trial randomised 3,445 patients with HF and an EF ≥45% to spironolactone or placebo, assessing a composite of cardiovascular death, survived cardiac arrest, and HF hospitalisation as the primary endpoint in a time-to-event analysis.⁴⁰⁾ The trial did not achieve statistical significance for the composite outcome (HR, 0.89; 95% CI, 0.77–1.04); however, in a secondary analysis, the risk of HF hospitalisation was significantly reduced with spironolactone (HR, 0.83; 95% CI, 0.69–0.99).⁴⁰⁾ Overall, TOPCAT was a neutral trial, yet its interpretation has been controversial. Patients were enrolled and treatment assignment stratified based on either a HF hospitalisation in the preceding year or brain natriuretic peptide (BNP)/NT-proBNP levels exceeding prespecified thresholds.⁴⁰⁾ In a prespecified interaction analysis, patients enrolled via BNP/NT-proBNP criteria had a significantly lower HR for the primary outcome than those enrolled based on prior hospitalisation (HR, 0.65 vs. 1.01; p-interaction=0.01).⁴⁰⁾ Further investigation revealed that patients enrolled on the basis of prior hospitalisation were, unexpectedly, at lower baseline risk for the primary outcome than those enrolled via natriuretic peptide criteria.⁴⁰⁾ The majority of patients from Russia and Georgia were enrolled in the hospitalisation stratum. Post hoc analyses suggested a benefit of spironolactone in patients from the Americas but not in those from Russia and Georgia, raising concerns regarding trial conduct and the validity of the HFpEF diagnosis in these countries.⁴⁰,⁸⁸,⁸⁹⁾

The ESC guidelines provide no recommendation for MRA therapy in HFpEF, whereas the AHA/ACC/HFSA guideline assigns a class IIb B-R recommendation, with caution due to the post hoc nature of the supporting evidence.⁵³,⁷²⁾ Although there is limited high-quality evidence for hard clinical outcomes, spironolactone has been shown to improve exercise capacity, cardiac structure, and function in HFpEF compared with placebo.⁹⁰,⁹¹⁾ A subanalysis of TOPCAT suggested greater efficacy of spironolactone at the lower end of the EF spectrum (inclusion with ≥45%), supporting IIb C/2b B-NR recommendations for patients with HFmrEF in both ESC and AHA/ACC/HFSA guidelines.⁵³,⁷²,⁹²⁾ Further clarification on the role of spironolactone in HFpEF and HFmrEF is expected from the ongoing SPIRRIT-HFpEF registry-based randomised controlled trial, with planned follow-up until the end of 2025, and the SPIRIT-HF trial which was recently terminated.⁹³,⁹⁴⁾

Notably, no randomised controlled trial of eplerenone has evaluated hard clinical outcomes in HFpEF or HFmrEF; however, available data suggest that its treatment effect is at least as favourable as that of spironolactone.⁹⁵⁾ The emerging evidence on non-steroidal MRAs is addressed in the following section.

EMERGING PHARMACOLOGICAL THERAPIES

Evidence from FINEARTS-HF

Recent advances in the field have emerged from non-steroidal MRAs. The non-steroidal MRA finerenone targets the MR with a potency comparable to spironolactone and a selectivity similar to eplerenone.⁹⁶⁾ Finerenone has also been shown to distribute more evenly between renal and cardiac tissue than steroidal MRAs, which tend to accumulate in the kidneys.³⁹,⁹⁶⁾ Beyond tissue distribution, non-steroidal MRAs exert distinct molecular effects at the receptor level. Finerenone induces a unique MR ligand-binding domain conformation, resulting in altered recruitment of transcriptional co-regulators compared with steroidal MRAs.⁹⁷⁾ This selective MR modulation leads to a more complete repression of pro-inflammatory and pro-fibrotic gene expression, while exerting less residual agonistic activity in epithelial tissues.⁹⁷⁾ These mechanistic differences provide a plausible biological rationale for the enhanced cardiorenal protection and improved tolerability profile observed with non-steroidal MRAs, including a lower risk of hyperkalaemia—a major barrier to MRA implementation.⁹⁷,⁹⁸,⁹⁹⁾

Two phase III trials initially evaluated finerenone in patients with CKD and type 2 diabetes. In FIDELIO-DKD, 5,734 patients were randomised to finerenone or placebo.¹⁰⁰⁾ Over a median follow-up of 2.6 years, the primary outcome—time to kidney failure, sustained ≥40% decline in eGFR, or death from renal causes—occurred in 17.8% of patients receiving finerenone vs 21.1% of those receiving placebo, accounting for a statistically significant reduction in risk by 18% (HR, 0.82; 95% CI, 0.73–0.93).¹⁰⁰⁾ The key secondary outcome—time to cardiovascular death, myocardial infarction, stroke, or HF hospitalisation—was also reduced with finerenone compared with placebo (HR, 0.86; 95% CI, 0.75–0.99).¹⁰⁰⁾ In the FIGARO-DKD trial, a broader spectrum of CKD with concomitant type 2 diabetes was assessed, with a focus on cardiovascular outcomes.¹⁰¹⁾ Among 7,437 patients randomised to finerenone or placebo, those receiving finerenone had a lower risk of the primary cardiovascular outcome—time to cardiovascular death, myocardial infarction, stroke, or HF hospitalisation (12.4% vs. 14.2%; HR, 0.87; 95% CI, 0.76–0.98)—over a median follow-up of 3.4 years, primarily driven by a reduced risk of HF hospitalisation.¹⁰¹⁾ FIDELITY, a prespecified individual patient data meta-analysis of FIDELIO-DKD and FIGARO-DKD, included 13,026 patients with a median follow-up of 3.0 years. The analysis confirmed that finerenone reduced the risk of the primary cardiovascular outcome (same as in FIGARO-DKD) compared with placebo (HR, 0.86; 95% CI, 0.78–0.95).¹⁰²⁾ A prespecified subanalysis of FIDELIO-DKD indicated that the cardiovascular benefit was not modified by a prior diagnosis of HF, present in 7.7% of participants.¹⁰³⁾ Consequently, the 2023 update of the 2021 ESC guidelines introduced a class I A indication for finerenone in the prevention of HF hospitalisations in patients with CKD and type 2 diabetes.²⁰⁾ The AHA/ACC/HFSA guidelines, however, do not currently provide any recommendation.⁷²⁾

The FINEARTS-HF trial was an event-driven, dedicated HF trial designed to further evaluate finerenone’s potential to improve mortality/morbidity. Across 37 countries, 6,001 patients with symptomatic HF and an EF ≥40% (i.e. HFmrEF or HFpEF) were randomised 1:1 to finerenone or placebo, both in addition to standard of care.²²⁾ The primary outcome was a composite of total worsening HF events (hospitalisations or urgent out-patient visits due to HF) and cardiovascular death.²²⁾ Over a median follow-up of 32 months, 1,083 primary events (60.1%) occurred among the 3,003 patients in the finerenone arm, compared with 719 events (39.9%) among 2,998 patients in the placebo arm, corresponding to a 16% significant reduction in event rate with finerenone (rate ratio, 0.84; 95% CI, 0.74–0.95).²²⁾ This result was consistent in an analysis of the primary composite outcome using a time-to-event approach.²²⁾ The cardiovascular benefit was primarily driven by a reduction in HF hospitalisations.²²⁾ Benefits were consistent across New York Heart Association classes, NT-proBNP levels, background therapies (renin-angiotensin-system inhibitors and SGLT2i), and comorbidities such as atrial fibrillation or type 2 diabetes.²²⁾ From a safety perspective, serious adverse events were overall less frequent with finerenone than placebo (38.7% vs. 40.5%).²²⁾ Hyperkalaemia defined as K >6.0 mmol/L occurred more often with finerenone (3.0% vs. 1.4%), though only 0.5% of patients in the finerenone arm required hospitalisation for hyperkalaemia (placebo: 0.2%), and no hyperkalaemia-related deaths occurred in either group.²²⁾ Hypokalaemia (i.e., <3.5 mmol/L) was less common with finerenone than placebo (4.4% vs. 9.7%).²²⁾ Both potassium imbalances are associated with increased mortality in HF—hypokalaemia with both short- and long-term mortality, hyperkalaemia primarily with short-term mortality.¹⁰⁴,¹⁰⁵⁾ The composite renal outcome (sustained ≥50% eGFR decrease, eGFR <15 mL/min/1.73 m², initiation of long-term dialysis, or kidney transplantation) occurred numerically more often with finerenone than with placebo (HR, 1.33; 95% CI, 0.94–1.89), but the difference was not statistically significant.²²⁾ Absolute event rates were low (finerenone: 75 events ≙2.5%; placebo: 55 events ≙1.8%).²²⁾

In an individual patient data meta-analysis of MRA trials in HF (HFrEF: RALES, EMPHASIS-HF; HFmrEF/HFpEF: TOPCAT, FINEARTS-HF), MRAs were associated with a reduction in the composite of HF hospitalisation or cardiovascular death across the EF spectrum, with greater efficacy observed in HFrEF compared with HFmrEF/HFpEF.¹⁰⁶⁾ No significant heterogeneity in treatment effect was detected across the HFmrEF/HFpEF trials, TOPCAT (both in the overall cohort and in a sensitivity analysis restricted to patients randomised in the Americas) and FINEARTS-HF, which leads to the question of whether MRAs might have a class effect in patients with HFpEF.¹⁰⁶⁾ Data from the ongoing trials SPIRRIT-HFpEF and REDEFINE-HF as well as the recently terminated SPIRIT-HF trial will eventually contribute to answer this question.⁹³,⁹⁴,¹⁰⁷⁾

Therefore, FINEARTS-HF provides novel evidence supporting non-steroidal MRA therapy in patients with HFmrEF or HFpEF, which may likely lead to a recommendation in the forthcoming ESC HF guidelines, scheduled for 2026 (Figure 1). Additional evidence on finerenone in hospitalised patients with HFmrEF or HFpEF is expected from the ongoing phase III REDEFINE-HF trial, with completion anticipated in April 2026.¹⁰⁷⁾

Advent of GLP-1 RAs

There is growing recognition that HFpEF is a heterogeneous syndrome, with increasing emphasis on identifying distinct phenotypes characterised by different clinical features, prognosis, and therapeutic needs. Obesity is present in 30–40% of patients with HFpEF and is associated with more severe symptoms, reduced quality of life, and poorer prognosis.⁴³⁾ The cardiometabolic profile is adversely affected by obesity—more accurately reflected by the waist-to-hip ratio than by body mass index—and patients with HFpEF who are obese tend to be younger than those without obesity.⁴³⁾ Lifestyle modification and weight reduction through exercise and dietary interventions are recommended, alongside optimal management of comorbidities.²⁰,⁴³,⁵³⁾

An emerging phenotype-specific pharmacotherapeutic approach in the management of patients with HFpEF and obesity will likely arise from the rapidly evolving class of GLP-1 RAs (Figure 1). In the STEP-HFpEF trials, semaglutide improved HF-related symptoms and reduced body weight in patients with HFpEF and obesity compared with placebo.¹⁰⁸,¹⁰⁹⁾ A greater reduction in NT-proBNP (change from baseline to week 52) with semaglutide than with placebo further suggested a specific benefit on HF.¹⁰⁸,¹⁰⁹⁾ A meta-analysis combining these two trials with the HFpEF subgroups from SELECT and FLOW also suggested a reduced risk of cardiovascular death or worsening HF events, primarily driven by the latter.¹¹⁰⁾ In SELECT, a prespecified subgroup of patients with HF (53% HFpEF, 31% HFrEF, remainder unclassified) had previously shown a numerically reduced risk of this outcome with semaglutide: worsening HF was more reduced in HFpEF, cardiovascular death more in HFrEF.¹¹¹⁾ In the FLOW trial, semaglutide reduced the composite renal outcome (death from renal or cardiovascular causes, sustained ≥50% eGFR decrease, eGFR <15 mL/min/1.73 m², initiation of long-term dialysis, or kidney transplantation) compared with placebo, with consistent effects in patients with and without prevalent HF.¹¹²⁾ Tirzepatide, a dual gastric inhibitory polypeptide and GLP-1 RA, was assessed in the SUMMIT trial—the first phase III hard outcome study of a GLP-1 RA in HF. In patients with HFpEF and obesity, tirzepatide reduced the composite risk of cardiovascular death or worsening HF events, an effect entirely driven by fewer worsening HF events with no reduction in mortality, as well as improved symptoms compared with placebo.¹¹³⁾

Although current guidelines do not yet recommend GLP-1 RAs for the treatment of HF, this emerging evidence from these trials may influence forthcoming recommendations.²⁰,⁵³,⁷²⁾ Even prior to such updates, treating physicians should consider GLP-1 RAs for patients with HF on the basis of approved indications, such as type 2 diabetes.

UNMET NEEDS AND FUTURE DIRECTIONS

The therapeutic landscape of HFpEF has changed markedly in recent years (Figure 2, Table 1).¹⁹,²¹,²²,⁴⁰,⁵⁷,⁵⁸,⁷¹,⁷³,⁷⁹,⁸⁰,⁸²,⁸⁴,⁸⁵,¹⁰⁸,¹⁰⁹,¹¹³,¹¹⁴⁾ Empagliflozin and dapagliflozin, supported by the EMPEROR-Preserved and DELIVER trials, provide the first disease-modifying therapies shown to improve prognosis in HFpEF.¹⁹,²¹⁾ While generally well tolerated, their implementation in routine clinical practice remains a key challenge.¹¹⁵⁾

Figure 2. Evolution of the pharmacotherapeutic landscape and future directions.

ARNi = angiotensin receptor-neprilysin inhibitor; SGLT2i = sodium-glucose cotransporter 2 inhibitor; MRA = mineralocorticoid receptor antagonist; GLP-1 RA = glucagon-like peptide-1 receptor agonist; HFpEF = heart failure with preserved ejection fraction; HFmrEF = heart failure with mildly reduced ejection fraction; RCT = randomised controlled trial.

Table 1. Landmark randomised controlled trials on pharmacotherapy in HFpEF.

Trial		Drug	Comparator	Sample size	EF inclusion criteria	Primary outcome(s)	Effect on primary outcome	NNT	Effect on mortality	Effect on quality of life (95% CI)	Adverse events (drug vs. comparator)	Follow-up*
SGLT2i
	EMPULSE58)	Empagliflozin	Placebo	530	Full EF spectrum (31% EF >40%)	Composite clinical benefit (i.e. all-cause death, HF events and time-to-first event, KCCQ change at 90 days)	Win ratio 1.36 (95% CI, 1.09–1.68)	N.A.	Cardiovascular death: 12.8% vs 18.5%; all-cause death: 4.2% vs 8.3%	Mean change in KCCQ at 90 days vs placebo: +4.45 points (95% CI, 0.32–8.59)	Volume depletion: 12.7% vs. 10.2%; acute kidney injury: 7.7% vs. 12.1%; urinary tract infection 4.2% vs. 6.4%; hypoglycaemia 1.9% vs. 1.5%; hypotension 1.2% vs. 1.5%; total serious adverse events 15.0% vs. 20.5%	90 days
	SOLOIST-WHF57)	Sotagliflozin	Placebo	1,222	Full EF spectrum (21% EF ≥50%)	Total number, including recurrent events, of cardiovascular death, hospitalisations for HF, and urgent HF visits	51.0 vs. 76.3 per 100 patient-years (HR, 0.67; 95% CI, 0.52–0.85)	7	Cardiovascular death: HR, 0.84 (95% CI, 0.58–1.22); all-cause death: HR, 0.82 (95% CI, 0.59–1.14)	Mean change in KCCQ at 4 months vs. placebo: +3.3 points (95% CI, 0.5–6.2)	Diarrhoea: 6.1% vs. 3.4%; hypotension: 6.0% vs. 4.6%; urinary tract infection: 4.8% vs. 5.1%; hypoglycaemia: 4.3% vs. 2.8%; hyperkalaemia (i.e., potassium >6.0 mmol/L): 4.3% vs. 5.1%; acute renal failure: 4.1% vs. 4.4%; serious adverse events: 38.8% vs. 41.1%	9 months
	EMPEROR-Preserved21)	Empagliflozin	Placebo	5,988	EF >40% (67% EF ≥50%)	Composite of cardiovascular death or hospitalisation for HF	6.9 vs. 8.7 per 100 patient-years (HR, 0.79; 95% CI, 0.69–0.90)	60	Cardiovascular death: HR, 0.91 (95% CI, 0.76–1.09); all-cause death: HR, 1.00 (95% CI, 0.87–1.15)	Mean change in KCCQ at 12 months vs. placebo: +1.32 points (95% CI, 0.45–2.19)	Acute renal failure: 12.1% vs. 12.8%; urinary tract infections: 9.9% vs. 8.1%; hypoglycaemia 4.3% vs. 4.5%; genital infections: 2.2% vs. 0.7%; ketoacidosis: 0.1% vs. 0.2%; serious adverse events: 47.9% vs. 51.6%	2.2 years
	DELIVER19)	Dapagliflozin	Placebo	6,263	EF >40% (66% EF ≥50%)	Composite of worsening HF (i.e., hospitalisation for HF or an urgent visit for HF) or cardiovascular death	7.8 vs. 9.6 per 100 patient-years (HR, 0.82; 95% CI, 0.73–0.92)	61	Cardiovascular death: HR, 0.88 (95% CI, 0.74–1.05); all-cause death: HR, 0.94 (95% CI, 0.83–1.07)	Mean change in KCCQ at 8 months vs. placebo: +2.4 points (95% CI, 1.5–3.4)	Acute myocardial infarction: 1.6% vs. 1.9%; acute renal failure: 1.5% vs. 1.6%; urinary tract infections: 1.0% vs. 1.0%; cellulitis: 1.0% vs. 0.6%; serious adverse events: 43.5% vs. 45.5%	2.3 years
	EMPEROR-Preserved + DELIVER pooled analysis71)	Dapagliflozin or empagliflozin	Placebo	12,251	EF >40%	Composite of cardiovascular death or first HF hospitalisation	14.5% vs. 17.8% (HR, 0.80; 95% CI, 0.73–0.87)	30	Cardiovascular death: HR, 0.88 (95% CI, 0.77–1.00)	No heterogeneity in treatment effect between trials (p=0.64)	Hypoglycaemia: 0.2–2.4% vs. 0.2–2.6%; amputation: 0.5–0.6% vs. 0.8%; ketoacidosis: 0.1% vs. 0.2%; serious adverse events: 43.5–47.9% vs. 45.5–51.6%	2.3 years
ARNi
	PARAGON-HF73)	Sacubitril/valsartan	Valsartan	4,822	EF ≥45 %	Composite of total (first and recurrent) hospitalisations for HF and cardiovascular death	12.8 vs. 14.6 per 100 patient-years (RR, 0.87; 95% CI, 0.75–1.01)	64	Cardiovascular death: HR, 0.95 (95% CI, 0.79–1.16); all-cause death: HR, 0.97 (95% CI, 0.84–1.13)	Mean change in KCCQ at 8 months vs valsartan: +1.0 points (95% CI, 0.0–2.1); at 6 months +0.52 points (95% CI, 0.93–1.97)	Hypotension: 23.2% vs. 17.0%; hyperkalaemia (i.e., potassium >5.5 mmol/L): 13.7% vs. 13.7%; acute renal failure: 5.6% vs. 6.6%; angioedema: 0.6% vs. 0.2%; serious adverse events: 58.9% vs. 58.9%	2.9 years
	PARADIGM-HF + PARAGON-HF (pooled analysis)77)	Sacubitril/valsartan	Enalapril in PARADIGM-HF; valsartan in PARAGON-HF	8,399	PARADIGM-HF EF eligibility ≤40%; PARAGON-HF EF eligibility ≥45% (30% EF >50%)	Time to cardiovascular death or first HF hospitalisation	Overall: 3,114 (23.6%) events (HR, 0.84; 95% CI, 0.78–0.90); EF >50%: 869 (21.4%) events (HR, 0.94; 95% CI, 0.82–1.08)	N.A.	Overall cardiovascular death: HR, 0.84 (95% CI, 0.76–0.92); cardiovascular death EF >50%: HR, 0.96 (95% CI, 0.77–1.20)	N.A.	Combined by EF for EF >52.5% to 62.5% vs. EF>32.5% to 42.5%: hypotension 13.9% vs. 18.0%; hyperkalaemia (i.e., potassium >6.0 mmol/L) 3.3% vs. 5.9%; creatinine ≥3.0 mg/dL 1.7% vs. 1.6%; overall serious adverse events: N.A.	2.6 years
	PARAGLIDE-HF79)	Sacubitril/valsartan	Valsartan	466	EF >40% (77% EF ≥50%)	Time-averaged proportional change in NT-proBNP (weeks 4–8)	Ratio of change 0.85 (95% CI, 0.73–0.999)	N.A.	Exposure-adjusted rate 100 patient-years for cardiovascular mortality: 6.8 (95% CI, 3.2–12.4) with sacubitril/valsartan vs 11.7 (95% CI, 6.9–18.5) with valsartan	N.A.	Symptomatic hypotension: 24.0% vs. 15.5%; hyperkalaemia (i.e., potassium >5.5 mmol/L): 19.3% vs. 18.5%; worsening renal function: 21.5% vs. 30.9%; angioedema: 0% vs. 0.4%; serious adverse: 44.0% vs. 44.2%	5.9 months
	PARAGLIDE-HF + PARAGON-HF (pooled analysis)80)	Sacubitril/valsartan	Valsartan	5,262	PARAGLIDE-HF EF eligibility >40%; PARAGON-HF EF eligibility ≥45% (~70% EF ≤60%)	Total (first + recurrent) worsening HF events (hospitalisations and urgent ambulatory visits for HF) and cardiovascular death	27.5 vs. 34.5 per 100 patient-years (RR, 0.78; 95% CI, 0.61–0.99) in recent WHF; 14.5 vs. 16.8 per 100 patient-years (RR, 0.86; 95% CI, 0.75–0.98) in overall	14 in recent WHF; 44 in overall	Cardiovascular death in recent WHF: HR, 0.81 (95% CI, 0.55–1.18); in overall: HR, 0.93 (95% CI, 0.77–1.12)	N.A.	Symptomatic hypotension: 20.9% vs. 16.8%; hyperkalaemia (i.e., potassium >5.5 mmol/L): 18.5% vs. 18.1%; worsening renal function: 20.9% vs. 27.1%; overall serious adverse events: N.A.	2.2 years
ARB
	CHARM-Preserved84)	Candesartan	Placebo	3,023	EF >40% (64% EF ≥50%)	Composite of cardiovascular death or HF hospitalisation	8.1 vs. 9.1 per 100 patient-years (HR, 0.89; 95% CI, 0.77–1.03)	109	Cardiovascular death: HR, 0.99 (95% CI, 0.80–1.22); all-cause death: HR, 1.10 (95% CI, 0.79–1.52)	N.A.	Increase in creatinine 4.8% vs. 2.4%; hyperkalaemia (i.e., potassium >6.0 mmol/L): 2.0% vs. 0.1%; hypotension: 2.4% vs. 1.1%; serious adverse events: 17.8% vs. 13.5%	3 years
	I-PRESERVE85)	Irbesartan	Placebo	4,128	EF ≥45%	Composite of all-cause death or hospitalisation for cardiovascular causes (i.e., HF, myocardial infarction, unstable angina, arrhythmia, or stroke)	10.0 vs. 10.5 per 100 patient-years (HR, 0.95; 95% CI, 0.86–1.05)	222	Cardiovascular death: HR, 1.01 (95% CI, 0.86–1.18); all-cause death: HR, 1.00 (95% CI, 0.88–1.14)	Mean change in MLHF score at 6 months vs placebo: −8 vs. −7 points	Doubling of creatinine: 6% vs. 4%; hyperkalaemia (i.e., potassium >6.0 mmol/L): 3% vs. 2%; hypotension: 3% vs. 3%; overall serious adverse events 33% vs. 34%	4.1 years
Beta-blocker
	SENIORS82)	Nebivolol	Placebo	2,128	EF ≤35% or >35% with prior HF hospitalisation (35% EF >35%)	Composite of all-cause mortality or cardiovascular hospitalisation (i.e., worsening HF, acute coronary syndromes, cerebrovascular causes, or other cardiovascular causes)	31.1% vs. 35.3% (HR, 0.86; 95% CI, 0.74–0.99)	24	Cardiovascular death: HR, 0.84 (95% CI, 0.66–1.07); all-cause death: HR, 0.88 (95% CI, 0.71–1.08)	N.A.	Bradycardia: 11.1% vs. 2.6%; hypotension: 7.7% vs. 7.2%; fatigue: 6.7% vs. 5.8%; overall serious adverse events: N.A.	1.7 years
MRA
	TOPCAT40)	Spironolactone	Placebo	1,722	EF ≥45%	Composite of death from cardiovascular causes, aborted cardiac arrest, or hospitalisation for HF	18.6% vs. 20.4% (HR, 0.89; 95% CI, 0.77–1.04)	56	Cardiovascular death: HR, 0.90 (95% CI, 0.73–1.12); all-cause death: HR, 0.91 (95% CI, 0.77–1.08)	N.A.	Hyperkalaemia (i.e., potassium >5.5 mmol/L): 18.7% vs. 9.1%; elevated creatinine: 10.2% vs. 7.0%; hypokalaemia (i.e., potassium <3.5 mmol/L): 16.2% vs. 22.9%; total serious adverse events 41.6 vs. 41.8 per 100 person-years	3.3 years
	FINEARTS-HF22)	Finerenone	Placebo	6,016	EF ≥45%	Composite of total worsening HF events (i.e., first or recurrent unplanned hospitalisation or urgent visit for HF) and death from cardiovascular causes	14.9 vs. 17.7 per 100 patient-years (RR, 0.84; 95% CI, 0.74–0.95)	42	Cardiovascular death: HR, 0.93 (95% CI, 0.78–1.11); all-cause death: HR, 0.93 (95% CI, 0.83–1.06)	KCCQ difference across 6, 9, and 12 months: 1.6 (95% CI, 0.8–2.3)	Systolic blood pressure <100 mmHg: 18.5% vs. 12.4%; hyperkalaemia (i.e., potassium >5.5 mmol/L): 14.3% vs. 6.9%; hypokalaemia (i.e., potassium <3.5 mmol/L): 4.4% vs. 9.7%; creatinine ≥3.0 mg/dL: 2.0% vs. 1.2%; serious adverse events 38.7% vs. 40.5%	2.7 years
GLP-1 RA
	STEP-HFpEF DM109)	Semaglutide	Placebo	616	EF ≥45%	Change in the KCCQ-CSS and in body weight from baseline to week 52	KCCQ-CSS difference +7.3, (95% CI, 4.1 to 10.4); body weight difference −6.4% (95% CI, −7.6 to −5.2)	N.A.	N.A.	See primary endpoint	Gastrointestinal disorders: 34.8% vs. 19.0%; cardiac disorders: 7.7% vs. 13.7%; infections: 4.8% vs. 8.8%; serious neoplasms: 2.9% vs. 3.3%; serious acute renal failure: 0.6% vs. 2.3%; serious gallbladder-related disorders: 1.0% vs. 0.7%; serious adverse events 17.7% vs. 28.8%	1 year
	STEP-HFpEF108)	Semaglutide	Placebo	529	EF ≥45%	Change in the KCCQ-CSS and in body weight from baseline to week 52	KCCQ-CSS difference +7.8, (95% CI, 4.8 to 10.9); body weight difference −10.7% (95% CI. −11.9 to −9.4)	N.A.	N.A.	See primary endpoint	Cardiac disorders: 3.0% vs. 12.4%; gastrointestinal disorders: 3.4% vs. 2.6%; infections: 2.7% vs. 6.4%; serious acute renal failure: 1.9% vs. 0.8%; serious neoplasms: 1.5% vs. 2.3%; serious adverse events 13.3% vs. 26.7%	1.0 years
	SUMMIT113)	Tirzepatide	Placebo	731	EF ≥50%	Composite of cardiovascular death or a worsening HF event (i.e., hospitalisation or urgent HF visit); change in the KCCQ-CSS from baseline to week 52	9.9% vs. 15.3% (HR, 0.62; 95% CI, 0.41–0.95); KCCQ-CSS difference +6.9 (95% CI, 3.3–10.6)	19†	Cardiovascular death: HR, 1.58 (95% CI, 0.52–4.83); all-cause death: HR, 1.25 (95% CI, 0.63–2.45)	See primary endpoint	Diarrhoea: 18.4% vs. 6.3%; nausea: 17.0% vs. 6.5%; constipation: 14.8% vs. 6.0%; decreased appetite: 10.4% vs. 1.6%; vomiting: 10.4% vs. 2.2%; urinary tract infection: 9.9% vs. 6.0%; atrial fibrillation: 6.3% vs. 3.3%; hypotension: 6.0% vs. 3.0%; upper abdominal pain: 5.5% vs. 1.9%; anaemia: 5.2% vs. 4.9%; total serious adverse events (occurring in ≥4 patients): 26.4% vs. 25.6%	2.0 years
Iron supplementation
	FAIR-HFpEF115)	Intravenous ferric carboxymaltose	Placebo	39	≥45%	Change in 6MWTD from baseline to week 24	45 vs. −8 m; difference in least-squares means +49 m (95% CI, 5–93)	N.A.	N.A.	KCCQ change from baseline to week 24 +0.49±14.81 vs. +12.41±11.53; between-group difference +6.5 (95% CI, −3.8 to 16.7)	(No. of events) cardiovascular: 16 vs. 30; respiratory: 5 vs. 5; musculoskeletal: 2 vs. 7; renal: 2 vs. 4; gastrointestinal: 2 vs. 4; dermatological: 2 vs. 3; infectiological: 3 vs. 0; neurological: 2 vs. 1; cognitive: 0 vs. 1; haematological: 0 vs. 1; metabolic: 0 vs. 1; nephrotoxic: 0 vs. 1; total serious adverse events: 5 vs. 19	10 months

HFpEF = heart failure with preserved ejection fraction; EF = ejection fraction; NNT = number needed to treat; CI = confidence interval; SGLT2i = sodium-glucose cotransporter 2 inhibitor; HF = heart failure; KCCQ = Kansas City Cardiomyopathy Questionnaire; N.A. = not available/applicable; HR = hazard ratio; ARNi = angiotensin receptor-neprilysin inhibitor; RR = rate ratio; NT-proBNP = N-terminal prohormone of brain natriuretic peptide; WHF = worsening heart failure; ARB = angiotensin receptor blocker; MLHF = Minnesota Living with Heart Failure; MRA = mineralocorticoid receptor antagonist; GLP-1 RA = glucagon-like peptide-1 receptor agonist; CSS = clinical summary score; 6MWTD = 6-minute walk test distance.

^*Median or mean duration; ^†Refers to the composite of cardiovascular death and worsening HF.

The availability of the non-steroidal MRA finerenone represents one of the most recent advances in the management of HFpEF. Based on the results of FINEARTS-HF, the U.S. Food and Drug Administration approved finerenone for patients with HFmrEF or HFpEF in July 2025.²²,¹¹⁶⁾ The forthcoming HF guidelines may likely incorporate corresponding treatment recommendations, although the review process by the European Medicines Agency is still ongoing. Further insights into finerenone in patients with HFpEF or HFmrEF hospitalised for HF (or recently discharged) will be provided by the REDEFINE-HF trial.¹⁰⁷⁾ The combination of finerenone with dapagliflozin is being investigated in the ongoing CONFIRMATION-HF trial with hospitalised (or recently discharged) patients across the EF spectrum.¹¹⁷⁾ CONFIRMATION-HF will assess the efficacy and safety of initiating both a non-steroidal MRA and an SGLT2i concurrently in HF. The CONFIDENCE trial recently demonstrated that simultaneous initiation of finerenone and empagliflozin in patients with CKD and type 2 diabetes reduced the albumin-to-creatinine ratio from baseline to 180 days compared with either agent alone.¹¹⁸⁾ Safety was shown by the absence of unexpected adverse events and a low incidence of treatment discontinuation (combination: 4.5%; finerenone: 3.4%; empagliflozin: 3.4%).¹¹⁸⁾

Another promising development concerns aldosterone synthase inhibitors, which act upstream of MRAs by inhibiting aldosterone biosynthesis, whereas MRAs such as finerenone act downstream by blocking aldosterone signalling at the receptor level of the target tissues. Several aldosterone synthase inhibitors are under investigation across all clinical phases, including balcinrenone in combination with dapagliflozin in the phase III BalanceD-HF trial.¹¹⁹⁾ Although therapies such as MRAs may attenuate fibrotic signalling, effective strategies to achieve fibrosis reversal remain an unmet need, and represent an important target for future mechanistic and translational research.

HFpEF is highly heterogeneous, with up to 80% of patients presenting with at least one cardiovascular, metabolic, pulmonary, renal, or geriatric comorbidity.⁴³⁾ Treatment of these conditions, guided by patient phenotyping, is recommended and can improve outcomes beyond the benefits of HF-specific therapy, which is indicated for all patients with HFpEF.²⁰,⁴³,⁵³,⁷²⁾ Iron deficiency, prevalent in up to 50% of patients with HFpEF, represents a targetable comorbidity identified through phenotyping.⁴³⁾ Given that exercise intolerance and impaired quality of life are dominant clinical features of HFpEF, interventions improving functional capacity may be of particular relevance in this population. In the FAIR-HFpEF trial, intravenous ferric carboxymaltose improved 6-minute walking distance compared with placebo, suggesting a potential benefit on exercise tolerance, although the study was not powered to draw firm conclusions regarding patient-reported quality-of-life or hard outcomes.¹¹⁴⁾ Beyond iron deficiency, patient phenotyping highlights important therapeutic gaps in specific HFpEF subgroups. Approximately 60–70% of patients with HFpEF are aged >65 years, with a substantial proportion presenting with frailty.⁴³⁾ While obesity is observed in around 30–40% of patients and represents a potential target for metabolic interventions such as GLP-1 RAs, up to 15–20% of patients exhibit a lean or cachectic phenotype, which is consistently associated with worse outcomes and may warrant specific therapeutic approaches.⁴³⁾

Timely echocardiography is essential for diagnosing HF and determining the EF phenotype, yet access is often delayed, postponing both diagnosis and treatment initiation.¹²⁰,¹²¹⁾ In the population-based REVOLUTION HF study, patients with suspected HF had a median wait of 40 days for echocardiography, with the highest HF hospitalisation risk within the first six weeks.¹²¹⁾ Risk stratification to prioritise patients for timely echocardiography is complicated by varying NT-proBNP thresholds according to setting, age, and factors such as obesity or renal function, particularly outside specialist care.¹²⁰,¹²¹⁾ Artificial intelligence tools may help integrate these parameters for individually tailored management. Meanwhile, REVOLUTION HF might support considering SGLT2i initiation before HF diagnosis is confirmed, as this drug class is safe in patients without HF, effective across all HF phenotypes, and indicated by comorbidities such as type 2 diabetes or CKD.¹²¹⁾

In summary, the management of HFpEF is evolving from a uniform, symptom-driven approach towards a more individualised strategy that combines disease-modifying therapies with targeted treatment of comorbidities and specific phenotypes. Despite recent advances, substantial unmet needs remain, particularly in frail patients and in addressing myocardial fibrosis. Future progress in HFpEF will depend not only on the development of novel therapies but also on improved phenotyping, timely diagnosis, and effective implementation of evidence-based treatments in routine clinical practice.