Heart Failure with Preserved Ejection Fraction: a Pharmacotherapeutic Update.

While guidelines for management of heart failure with reduced ejection fraction (HFrEF) are consensual and have led to improved survival, treatment options for heart failure with preserved ejection fraction (HFpEF) remain limited and aim primarily for symptom relief and improvement of quality of life. Due to the shortage of therapeutic options, several drugs have been investigated in multiple clinical trials. The majority of these trials have reported disappointing results and have suggested that HFpEF might not be as simply described by ejection fraction as previously though. In fact, HFpEF is a complex clinical syndrome with various comorbidities and overlapping distinct phenotypes that could benefit from personalized therapeutic approaches. This review summarizes the results from the most recent phase III clinical trials for HFpEF and the most promising drugs arising from phase II trials as well as the various challenges that are currently holding back the development of new pharmacotherapeutic options for these patients.

Introduction

Although initially believed to be less severe than heart failure with reduced ejection fraction (HFrEF), studies show that heart failure with preserved ejection fraction (HFpEF) prevalence has increased and accounts for as much as 50% of heart failure (HF) cases [1]. With an increasing incidence and prevalence of the comorbidities closely related to HFpEF, such as hypertension [2], coronary artery disease [3], obesity [4], diabetes mellitus [5], chronic obstructive pulmonary disease [6], and chronic kidney disease [7], it is expected that the prevalence of HFpEF will further escalate. The recent SARS-CoV-2 pandemic has also shown to have some association with HFpEF by either causing, unmasking, or exacerbating existing HFpEF [8] and could overall contribute to the increasing prevalence of this clinical syndrome. Furthermore, in addition to the increasing number of patients, it is expected that hospitalization and mortality will increase its economic burden in the world’s health services. The global economic burden of HF has been estimated at US$108 billion per annum [9] with the most significant costs deriving from patient hospitalization [10]. A comprehensive systematic review recently conducted by Clark et al. found that HFpEF hospitalizations represent about 60% of total HF hospitalization costs and that the high prevalence of comorbidities in this patient population further aggravates its economic burden [11]. Despite its increasing prevalence and economic burden, treatment options for HFpEF are limited, and because patients are often elderly, highly symptomatic and have decreased quality of life, the aim of therapy is primarily symptom relief and improvement of quality of life [12, 13]. Diuretics are often used to improve congestion if present. However, therapy usually prescribed to treat other cardiovascular comorbidities (such as beta-blockers, mineralocorticoid receptor antagonists, angiotensin receptor blockers, or angiotensin-converting enzyme inhibitors) has lacking or inconsistent evidence on the improvement of symptoms or reduction of mortality in HFpEF [13]. As this pathology has a relatively high morbimortality [13], there is an urgent need for effective therapies.

Several pathophysiologic mechanisms lead to increased left ventricle (LV) end-diastolic pressure, causing HF symptoms [14]. Diastolic dysfunction in HFpEF patients results primarily from myocardial stiffness, a process largely regulated by the extracellular matrix, by both its composition and structure [15, 16], and cardiomyocytes, through the prolongation of Ca2+ transients [17, 18]. Moreover, various studies have found that alterations in titin are involved in the increased passive stiffness of the failing myocardium [19-21]. The strong association of HFpEF with chronic comorbidities also underlies a pathophysiological paradigm based on increased proinflammatory state and microvascular endothelial dysfunction contributing to impaired myocardial relaxation and compliance [22].

In recent years, several drugs have undergone phase II and phase III clinical trials for their potential as a novel pharmacological option for patients with HFpEF (Tables 1 and 2). A great percentage of these studies reported either disappointing results or no results at all. For the purposes of this review, we will explore the most recent findings from phase III clinical trials for HFpEF patients and the drugs that upon phase II trials showed most promising results as well as the various challenges that are currently holding back the development of new pharmacotherapeutic options for these patients.

Table 1

Phase III clinical trials

Acronym	Participants	Identifier	Intervention	Posology	Administration	Expected date for primary completion	Outcomes
PARAGON	4822	NCT01920711	Sacubitril-valsartan	100 to 200 mg b.i.d. + 80 mg b.i.d.	P.O.	June 2019	Non-significant lower rate of total hospitalizations for HF and death from CV causes. Significant NYHA class improvement and a reduction in the worsening of renal function [25]
PARALLAX	2572	NCT03066804	Sacubitril-valsartan	24/26 mg to 97/103 mg b.i.d.	P.O.	October 2019	No improvement in the 6MWTD, but significant reduction in NT-pro-BNP. It significantly reduced the decline in renal function and the risk for HF hospitalizations by 50% [34]
PRISTINE	60	NCT04128891	Sacubitril-valsartan	49/51 mg to 97/103 mg b.i.d.	P.O.	February 2022	Withdrawn (funding not approved)
PARAGLIDE	800	NCT03988634	Sacubitril-valsartan	24/26 mg to 97/103 mg b.i.d.	P.O.	March 2022	Not reported
PERSPECTIVE	592	NCT02884206	Sacubitril-valsartan	100 to 200 mg b.i.d. + 40 to 160 mg b.i.d.	P.O.	March 2022	Not reported
ARNI-MEMS	14	NCT04753112	Sacubitril-valsartan	97/103 mg b.i.d.	P.O.	October 2022	Not reported
TOPCAT	3445	NCT00094302	Spironolactone	15 to 45 mg o.d.	P.O.	June 2013	No significant decrease in CV mortality, aborted cardiac arrest or hospitalization [49]
SPIRRIT	3200	NCT02901184	Spironolactone	25 to 50 mg o.d.	P.O.	December 2021	Not reported
SPIRIT-HF	1300	NCT04727073	Spironolactone	25 to 50 mg o.d.	P.O.	December 2024	Not reported
FINEARTS-HF	5500	NCT04435626	Finerenone	10 to 20 mg or 20 to 40 mg o.d.	P.O.	March 2024	Not reported
DETERMINE-preserved	504	NCT03877224	Dapagliflozin	10 mg o.d.	P.O.	July 2021	Not reported
	648	NCT03794518	Pioglitazone + dapagliflozin	15 mg + 10 mg		September 2021	Not reported
DELIVER	6263	NCT03619213	Dapagliflozin	10 mg o.d.	P.O.	November 2021	Not reported
EMPERIAL-preserved	315	NCT03448406	Empagliflozin	10 mg o.d.	P.O.	October 2019	No improvement in the 6MWTD but improved the KCCQ-TSS score by at least 8 points [52]
EMPEROR-preserved	5988	NCT03057951	Empagliflozin	10 mg o.d.	P.O.	April 2021	Reduced mortality and hospitalization rates [53]
	200	IRCT2019012 2042450N2	Empagliflozin	10 mg o.d.	P.O.		Not reported
	8	NCT05139472	Empagliflozin	10 mg o.d.	P.O.	December 2022	Not reported
	71	NCT01411735	Enalapril	2.5 to 10 mg b.i.d.	P.O.	June 2004	Not reported
ULTIMATE-HFpEF	52	NCT01599117	Udenafil	50 to 100 mg b.i.d.	P.O.	April 2013	Not reported
	52	NCT01726049	Sildenafil	20 to 60 mg t.i.d.	P.O.	September 2014	No effects on hemodynamic parameters, as well as no improvements in cardiac structure or function, cardiopulmonary exercise testing, laboratory parameters, or quality of life [62, 63]
PASSION	372	DRKS00014595	Tadalafil	20 to 40 mg			Not reported
SOUTHPAW	84	NCT03037580	Treprostinil	0.125 to 6 mg t.i.d.	P.O.	December 2019	Terminated by sponsor
TDE-HF-302	48	NCT03043651	Treprostinil	0.125 to 6 mg t.i.d.	P.O.	March 2020	Terminated by sponsor
PREFER-HF	72	NCT03833336	Ferric carboxymaltose	500 to 1000 mg	I.V.	December 2019	Not reported
			Ferroglycine sulfate	100 mg	P.O.
			Sucrosomial iron	30 mg	P.O.
STEP-HFpEF	516	NCT04788511	Semaglutide	0.25 to 2.4 mg o.w.	S.C.	March 2023	Not reported
STEP HFpEF DM	610	NCT04916470	Semaglutide	0.25 to 2.4 mg o.w.	S.C.	June 2023	Not reported
SUMMIT	700	NCT04847557	Tirzepatide		S.C.	November 2023	Not reported
COLpEF	426	NCT04857931	Colchicine	0.5 mg o.d. or b.i.d.	P.O.	July 2024	Not reported
BEAT HFpEF	30	NCT02885636	Albuterol	2.5 mg	INH	September 2017	Improves pulmonary vascular load during exercise, CO, RV-PA coupling, and left heart filling without increasing pulmonary capillary hydrostatic pressures [86]
EDIFY	179	EudraCT 2012–002,742-20	Ivabradine	2.5 to 7.5 mg b.i.d.	P.O.	February 2016	No improvement in: echo-Doppler E/e′ ratio, 6MWTD, and plasma NT-proBNP concentration [87]
	40	jRCTs051200059	Ivabradine				Not reported

6MWTD 6-min walk test distance, b.i.d. twice a day, CO cardiac output, CV cardiovascular, E/e’ ratio between early mitral inflow velocity and early mitral annular diastolic velocity, HF heart failure, I.V. intravenous injection, INH inhaled, KCCQ-TSS Kansas City Cardiomyopathy Questionnaire Total Symptom Score, NT-pro-BNP N terminal pro-brain natriuretic peptide, NYHA New York Heart Association functional class, o.d. once a day, o.w. once a week, P.O. oral treatment, PA pulmonary artery, RV right ventricle, S.C. subcutaneous injection, t.i.d. three times a day; fields left blank no available data

Table.2

Phase II clinical trials

Acronym	Participants	Identifier	Intervention	Posology	Administration	Expected date for primary completion	Outcomes
SOCRATES-PRESERVED	477	NCT01951638	Vericiguat	1.25 to 10 mg o.d.	P.O.	August 2015	Improved patients KCCQ physical limitation score [94]
VITALITY-HFpEF	789	NCT03547583	Vericiguat	2.5 to 15 mg o.d.	P.O.	August 2019	Did not improve the KCCQ physical limitation score or 6MWTD [93]
CAPACITY-HFpEF	196	NCT03254485	IW-1973	40 mg o.d.	P.O.	August 2019	No effect in peak VO2, biomarker levels, or echocardiographic parameters [111]
DYNAMIC	118	NCT02744339	Riociguat	1.5 mg t.i.d.		August 2020	Not reported
ERADICATE-HF	36	NCT03416270	Ertugliflozin	15 mg o.d.	P.O.	March 2021	Not reported
STADIA-HFpEF	26	NCT04475042	Dapagliflozin	10 mg o.d.	P.O.	November 2021	Not reported
CAMEO-DAPA	51	NCT04730947	Dapagliflozin	10 mg o.d.	P.O.	January 2023	Not reported
	28	NCT01932606	Sodium nitrite	50 μg/kg/min for 5 min	Infusion during cardiac catheterization procedure	October 2014	Significantly improved exercise PCWP, resulting in a reduction in left heart filling pressures with exercise. Associated with increased LV stroke work with exercise. [102]
	26	NCT02262078	Sodium nitrite	90 mg	INH	December 2015	Reduced PCWP both at rest and during exercise [103]
ONOH	15	NCT02918552	Sodium nitrite	20 or 40 mg t.i.d.	P.O.	December 2018	Unpublished
INABLE	100	NCT02713126	Sodium nitrite	40 mg t.i.d.	P.O.	March 2022	Not reported
	26	NCT03015402	Sodium nitrite	40 mg t.i.d.	P.O.	March 2022	Not reported
NEAT-HFpEF	110	NCT02053493	Isosorbide mononitrate	30 to 120 mg o.d.	P.O.	February 2015	Did not improve daily activity level, 6MWTD, post-walk dyspnea score, quality of life scores, or NT-proBNP levels Dose-dependent decrease in daily activity levels [112, 113]
KNO3CK OUT HFPEF	76	NCT02840799	Potassium nitrate	6 mmol t.i.d.	P.O.	November 2021	Not reported
MPMA	53	NCT04913805	Potassium nitrate	6 mmol t.i.d.	P.O.	September 2026	Not reported
			KNO3 + PLC + NR	6 mmol t.i.d. + 1000 mg b.i.d. + 300 mg t.i.d.
	54	NCT01516346	Isosorbide dinitrate	20 or 40 mg t.i.d.	P.O.	March 2018	Did not reduce reflection magnitude or improve LV remodeling. Very poorly tolerated [114]
			Isosorbide dinitrate + hydralazine	20 or 40 mg t.i.d. + 37.5 or 75 mg t.i.d.
	17	NCT01919177	Nitrate-rich beetroot juice	140 mL (12 mmol of NO−3)	P.O.	September 2014	No changes in peak exercise efficiency. A single dose prior to exercise significantly improves peak VO2 and CO at peak exercise and reduces SVR [95]
STOP-EF	50	NCT02949531	O2	21 to 40%	INH	February 2017	Resulted in a small increase in exercise time but had no effect on peak workload [115]
D-HART2	31	NCT02173548	Anakinra	100 mg o.d.	S.C.	April 2017	Inhibited systemic inflammatory response but failed to improve aerobic exercise capacity or ventilation efficiency [116]
PANACHE	305	NCT03098979	Neladenoson bialanate	5 to 40 mg o.d.	P.O.	May 2018	No significant improvement in 6MWTD, KCCQ overall score, physical activity level, or cardiac biomarkers [117]
RALI-DHF	20	NCT01163734	Ranolazine	2 bolus + 1000 mg b.i.d.	I.V. + P.O.	February 2011	Resulted in modest improvements in LV end-diastolic pressure, PCWP, and mPAP but decreased CO and SV [107]
SERENADE	143	NCT03153111	Macitentan	10 mg o.d.	P.O.	March 2021	Not reported
SERENADE OL	90	NCT03714815	Macitentan	10 mg o.d.	P.O.	May 2026	Not reported
AMETHYST	435	NCT04327024	Verinurad + allopurinol	3 to 24 mg + 100 to 300 mg	P.O.	November 2022	Not reported
	55	NCT04318145	PL-3994		I.V.	October 2021	Not reported
EMBARK-HFpEF	35	NCT04766892	Mavacamten		P.O.	May 2022	Not reported
	122	NCT04317339	Zhigancao Tang granule	200 mL b.i.d.	P.O.	March 2022	Not reported
	70	NCT00839228	Perhexiline	100 mg b.i.d.	P.O.	February 2014	Not reported
FAIR-HFpEF	200	NCT03074591	Ferric carboxymaltose	50 mg/mL	PAR	July 2020	Not reported
	56	NCT00286182	Erythropoietin alpha	7500U o.w.	S.C.	November 2012	Did not result in significant changes in LV structure nor function. No effects were seen in submaximal exercise capacity or in quality of life [118]
	46	NCT02814097	Elamipretide	40 mg o.d.	S.C.	May 2017	Not reported
CELLpEF	30	NCT02923609	CD34 + cell therapy		Transendocardial	January 2022	Not reported
PARAMOUNT	307	NCT00887588	LCZ696	50 to 200 mg b.i.d.	P.O.	December 2011	LCZ696 reduced NT-proBNP levels and LA size to a greater extent that valsartan. NYHA class improved significantly in patients on LCZ696 [35]
			Valsartan	40 to 160 mg b.i.d.
	60	NCT03928158	LCZ696	50 to 200 mg b.i.d.	P.O.	November 2020	Not reported
			Valsartan	40 to 160 mg b.i.d.
ENCHANTMENT	50	NCT04153136	Sacubitril-valsartan	49/51 mg b.i.d.	P.O.	June 2024	Not reported
ARNICFH	60	NCT05089539	Sacubitril-valsartan	100 mg b.i.d.	P.O.	February 2022	Not reported
DOT3HF-HFpEF	28	NCT04111536	Liothyronine	2.5 to 12.5 μg t.i.d.	P.O.	April 2023	Not reported
HELP	38	NCT03541603	Levosimendan	50 μg/min solution o.w.	I.V.	April 2020	Significantly decreased PCWP, CVP and submaximal exercise capacity [110]
	36	NCT03624010	Levosimendan	50 μg/min solution	I.V.	February 2024	Not reported
SATELLITE	41	NCT03756285	AZD4831		P.O.	May 2020	Terminated
	30	NCT03611153	AZD4831	30 mg	P.O.	April 2022	Not reported
ENDEAVOR	1485	NCT04986202	AZD4831	2.5 to 5 mg	P.O.	September 2024	Not reported
	20	NCT04633460	Ketone ester		P.O.	July 2022	Not reported
AVANTI	482	NCT03901729	Pecavaptan	30 mg o.d.	P.O.	April 2021	Not reported
			Furosemide	80 mg o.d.
	10	NCT03629340	Metformin	500 to 1000 mg b.i.d.	P.O.	September 2023	Not reported
	20	NCT05093959	Metformin	1500 mg o.d.	P.O.	December 2023	Not reported
PIROUETTE	129	NCT02932566	Pirfenidone	801 mg t.i.d.	P.O.	November 2019	Not reported
	102	NCT03882710	Metoprolol XR	25 to 100 mg o.d.	P.O.	October 2012	Not reported
	60	NCT02779634	CoQ10	100 mg t.i.d.	P.O.	January 2018	Not reported
	153	NCT03133793	CoQ10	300 mg b.i.d.	P.O.	March 2021	Not reported
			D-ribose	15 g o.d.	P.O.
CADENCE	180	NCT04945460	Sotatercept	0.3 to 0.7 mg/kg Q3W	S.C.	August 2023	Not reported
	225	NCT04944706	Qishen Yiqi dripping pills		P.O.	July 2023	Not reported
	20	ACTRN12614000727640	Milrinone	50 μg/kg for 10 min	I.V.		Decreased RA pressure, mPAP, and PCWP during exercise but showed no effect on the rate of isovolumic relaxation, LV stiffness, or EDPVR [64, 65]
	150	ChiCTR2000030769	Neucardin	0.8 μg/kg/ for 8 h/day or 0.27 μg/kg/ t.i.d.	S.C.	June 2023	Not reported
BRILLIANT	150	jRCT1031210030	Beta-blocker withdrawal				Not reported
	25	NCT05126836	Cilostazol	100 mg b.i.d.	P.O.	June 2022	Not reported
SAK HFpEF	53	NCT05138575	Empagliflozin + KCl	10 mg o.d. + 6 mmol t.i.d.	P.O.	September 2026	Not reported
			Empagliflozin + KNO3	10 mg o.d. + 6 mmol t.i.d.
			KCl	6 mmol t.i.d.
	296	NCT02599480	Mirabegron	50 mg o.d.	P.O.	August 2022	Not reported

6MWTD 6-min walk test distance, b.i.d. twice a day, CO cardiac output, CVP central venous pressure, EDPVR end-diastolic pressure–volume relationship, I.V. intravenous injection, INH inhaled, KCCQ Kansas City Cardiomyopathy Questionnaire, KCl potassium chloride, KNO potassium nitrate, LA left atrium, LV left ventricle, mPAP mean pulmonary artery pressure, NR nicotinamide riboside, NT-pro-BNP N terminal pro-brain natriuretic peptide, NYHA New York Heart Association functional class, o.d. once a day, o.w. once a week, P.O. oral treatment, PAR parenteral, PCWP pulmonary capillary wedge pressure, PLC propionyl-L-carnitine, Q3W every 3 weeks, RA right atrium, S.C. subcutaneous injection, SV stroke volume, SVR systemic vascular resistance, t.i.d. three times a day, U units, VO oxygen uptake, fields left blank no available data

Current Therapeutical Challenges

Despite recent developments in HFpEF pharmacological options, there still are no therapies proven to reduce mortality in this cohort of patients. This contrasts with HFrEF, for which there is panoply of pharmacological weapons in our arsenal, with some of the “big ones” (such as sacubitril-valsartan, or LCZ696) being recent discoveries [23].

One of the main possible arguments for some trials’ disappointing results is that HFpEF might not be as simply described as previously thought—just by the ejection fraction (EF); it is a complex syndrome with associated comorbidities and overlapping different phenotypes. In turn, this pushes us to think that, maybe, it is not the drugs that are ineffective, but it is the enormous heterogeneity of the patient population that predisposes the clinical trials to disappointing results [24]. Several post hoc analyses of the recent sacubitril-valsartan PARAGON trial (NCT01920711) point in this exact direction—despite the trial’s failure to meet its primary endpoint of reducing the number of composite events of cardiovascular death and total hospitalizations related to HF [25], there are studies showing significant results comparing either other relevant endpoints or other patient subgroups from the trial. For example, regarding the timing when the drug is given after a hospitalization, it seems that there is an amplification in the relative and absolute benefits of sacubitril-valsartan compared to only valsartan when the drug is administered early after hospitalization [26]. Also, when we compare the drug’s effect across the EF spectrum, we see a clear trend towards the reduction of it’s effect in preventing first HF hospitalization or cardiovascular death as the EF increases [27]. Moreover, in women, the drug is effective at higher EF than in men [27]. All this shows that not only can we plan the therapy based on EF, but also based on the comorbidities and characteristics of each patient.

The regular empirical use of beta-blockers for HFpEF is a good example of why there was a need to create more specific subgroups regarding the EF of the patients. There are only two clinical trials that studied the effects of beta-blockers in HFpEF patients: the SENIORS trial [28] and the J-DHF trial [29]. Regarding the first, although the results looked promising when using an EF cutoff of > 35%, in post hoc analysis, the subgroup with EF > 50% showed no benefits [30]. It must be said, as a sidebar, that as these trials were not designed to study the effects of beta-blockers specifically in the HFpEF population, therefore these results cannot lead to definitive and strong conclusions about the effects of beta-blockers in this population. This example is one of the many that led the European Society of Cardiology to create a new HF subgroup in 2016—heart failure with mid-range ejection fraction, in which the EF ranges between 40 and 50% [13]. This subgroup includes 14% of all HF patients [31], with an overlap of HFrEF and HFpEF phenotypes, but showing more similarities to the HFpEF subgroup [32]. This allows us to better design trials and guidelines and to better tailor each patient’s therapy.

Main Pharmacological Therapies

Renin–Angiotensin–Aldosterone System Inhibitors

Angiotensin Receptor-Neprilysin Inhibitor

Sacubitril-valsartan has just become the first drug to be indicated by the Food and Drug Association for the treatment of HFpEF. As mentioned before, the PARAGON trial showed a narrow miss in achieving its primary endpoint (risk ratio 0.87, 95% CI 0.75–1.01, p = 0.06) and showed significant protective results for the subgroup of patients with an EF below 57% (risk ratio 0.78, 95% CI (0.64–0.95)) [25]. Some authors argue that these results do not point towards the effectiveness of sacubitril-valsartan in HFpEF, but towards a need for change in the cutoffs between HFrEF and HFpEF, as this trial showed its best results in the “best EF for HFrEF”/ “worst EF for HFpEF” subgroups [33]. The treatment also showed better benefit in women (risk ratio 0.73, 95% CI (0.59–0.90)), who represent a high proportion of patients with HFpEF, than in men (risk ratio 1.03, 95% CI (0.85–1.25)). Secondary outcomes in the PARAGON trial were defined as the change in the clinical summary score on the Kansas City Cardiomyopathy Questionnaire (KCCQ), change in New York Heart Association (NYHA) functional class, first occurrence of a decline in renal function, and death from any cause. Sacubitril-valsartan showed significant benefits in changes in patients’ NYHA functional class and renal function, when compared to valsartan alone. During randomized treatment, sacubitril-valsartan was associated with higher incidence of hypotension and angioedema but with lower incidence of elevated serum creatinine and potassium levels than valsartan [25].

The PARALLAX clinical trial (NCT03066804) studied the effects of sacubitril-valsartan versus optimal individualized background therapy, which could be either an angiotensin II receptor blocker, an angiotensin-converting enzyme inhibitor, or a placebo [34]. This trial showed a significant reduction of NT-pro-BNP levels after 12 weeks of treatment; however, it failed to show improvement in the 6-min walk test distance (6MWTD). Furthermore, the results included a significant decrease in renal function worsening and a reduced risk for HF hospitalization by 50% [34]. The patients enrolled in this study were selected by having a KCCQ score lower than 75, showing an impacted quality of life; however, after 24 weeks of treatment, there were no differences in the KCCQ score [34].

There are several more phase III trials currently happening (or finished but still without published results).

The PRISTINE-HF trial (NCT04128891) enrolled 60 patients. It has a primary endpoint of showing differences in the microvascular function and in cardiac ischemia, with more clinical secondary endpoints (such as changes in the NYHA functional class, differences in the 6MWTD, cardiac mortality, and HF-related hospitalizations).

The PARAGLIDE-HF trial (NCT03988634) focuses on showing differences in the NT-pro-BNP levels in the group treated with sacubitril-valsartan, compared to patients only taking valsartan.

The PERSPECTIVE trial’s (NCT02884206) objective is to show differences in the cognitive function of patients with HFpEF treated with sacubitril-valsartan, using the CogState Global Cognitive Composite Score as an indicator of cognitive function and comparing with HFpEF patients taking only valsartan.

Recently, 14 patients with HFpEF and pulmonary hypertension (PH), taking sacubitril-valsartan and implanted with an CardioMEMS HF System—a device implanted in the pulmonary artery which continuously measures the mean pulmonary artery pressure (mPAP), offering real-time data on this parameter—were enrolled in the ARNIMEMS clinical trial (NCT04753112), which aims to enlighten us on the real-time effects of this drug on mPAP, blood biomarkers, and both functionality and quality of life of the patients.

As is evident by the current existence of phase III clinical trials, phase II trials for sacubitril-valsartan showed remarkably promising results. The PARAMOUNT (NCT00887588) trial showed significant reduction in NT-proBNP blood concentration when comparing the use of sacubitril-valsartan with valsartan-only treated patients, as well as reduction in the left atrium size and greater improvement in the patients NYHA functional class [35]. Thus, this trial provided us with the preliminary results for the efficacy and the safety of the drug in patients with HFpEF.

Moreover, there is an ongoing phase II clinical trial, ENCHANTMENT-HIV (NCT04153136), evaluating whether this medication could be useful to reduce HIV-related HFpEF, in patients between 40 and 70 years old with controlled HIV. Overall, this study aims to investigate the effect of sacubitril-valsartan on measures of heart disease related to inflammation, structure, and function in HIV, using the primary outcome measures of myocardial inflammation/fibrosis and left atrial volume index.

Angiotensin-Converting Enzyme Inhibitors

Since some evidence has suggested a potential role for angiotensin II in the pathophysiology of exercise intolerance in HFpEF patients [36-40], angiotensin antagonism has been hypothesized to be of interest in targeting exercise intolerance in older patients with HFpEF. The first study evaluating angiotensin-converting enzyme inhibitors, the PEP-CHF trial, evaluated perindopril’s effects on elderly patients with EF between 40 and 50% [41]. The trial failed to achieve its primary endpoint (composite of all-cause mortality or unplanned heart failure related hospitalization) for several reasons, including low event rate and large number of patients stopping assigned treatment after 1 year. The reduction in hospitalizations for HF and the reduction in primary endpoint approached conventional levels of statistical significance over the first year of follow-up. However, the trial did not show a statistical significant benefit of the drug on long-term morbidity and mortality. Enalapril was recently evaluated for its effect on exercise capacity and aortic distensibility in patients presenting with diastolic dysfunction (EF > 50%) (NCT01411735). Unfortunately, this study showed that enalapril administration failed to meet the defined endpoints, with no improvement seen on exercise capacity, aortic distensibility, or LV mass and volume after 12-month treatment [42].

Angiotensin Receptor Blockers

Similarly to what was hypothesized with angiotensin-converting enzyme inhibitors, angiotensin receptor blockers have been thought to be of benefit in patients with HFpEF. The CHARM-preserved study (NCT00634712) evaluated the effects of candesartan on the composite outcome of cardiovascular mortality or admission to hospital for worsening HF [43]. Though the trial found a trend towards fewer cardiovascular outcomes, favouring candesartan, it was moderate and of borderline significance. Even so, the numbers of individuals admitted one or more times for HF were reduced, reinforcing that candesartan might be of some benefit in this population [43]. The I-preserved trial (NCT00095238) evaluated irbesartan’s effect on [44] the composite outcome of death from any cause or hospitalization for a protocol-specified cardiovascular cause in patients with HFpEF [44]. The trial found that treatment with irbesartan did not reduce the risk of death or hospitalization for cardiovascular causes, nor did it improve any of the secondary clinical outcomes, such as patient quality of life. Further studies found similar results in which the use of angiotensin receptor blockers did not significantly improve patients outcomes [45-47]. A recent systematic review and meta-analysis of both randomized trials and observational studies found that both angiotensin-converting enzyme inhibitor and angiotensin receptor blockers were associated with a modest, but statistically significant, reduction in all-cause mortality in HFpEF patients [48]. However, in randomized trials alone, this effect was not seen. The results from this meta-analysis suggest that it may be important to further investigate these pharmacological classes in patients with HFpEF [48].

Aldosterone Receptor Antagonists

The first major clinical trial using spironolactone for HFpEF was the TOPCAT trial (NCT00094302). While it did fail to meet the primary composite outcome of cardiovascular mortality, aborted cardiac arrest, or HF-related hospitalizations, the spironolactone arm showed a significantly lower rate of hospitalizations for the management of a HF exacerbation (risk ratio 0.83, 95% CI (0.69–0.99), p = 0.042), despite not having an effect in the number of all-cause hospitalization [49]. Furthermore, patients taking spironolactone showed significantly greater incidences for hyperkalemia and increased serum creatinine levels [49]. However, some important differences in baseline characteristics were noted. The majority of the patients enrolled from Russia and Georgia had been hospitalized for HF in the 12 months prior to randomization, whereas patients enrolled from the USA, Canada, Argentina, and Brazil were more evenly balanced between hospitalized and non-hospitalized strata. Indeed, there was a marked regional variation in event rates, with patients on placebo group who had been enrolled in Russia or Georgia having a much lower likelihood of a primary outcome event than those enrolled in the Americas [49]. The discrepancy in event rates in the placebo group could have contributed to the observed treatment benefit in the Americas that was not seen in Russia or Georgia. In a post hoc subgroup analysis, the potential benefit of spironolactone with respect to the primary outcome was greatest in patients at the lower end of the EF spectrum (EF < 50%), most prominently found in patients enrolled in the Americas [50]. Treatment-EF interaction for the primary outcome was somewhat more pronounced in men (p = 0.01) than in females (p > 0.80) [50]. Given the FDA’s latest endorsement of sacubitril-valsartan in HFpEF, this could be an important finding.

The suggestion that spironolactone was effective in HFpEF was the basis for two ongoing phase III trials—the SPIRRIT-HF (NCT02901184) and the SPIRIT-HF trials (NCT04727073). Adding to these, we are waiting for the publication of the FINEARTS-HF clinical trial (NCT04435626) results, studying the effects of finerenone in HFpEF, a drug that showed robust results in the ARTS-HF phase IIb trial, not only in regard to safety, but also in the clinical outcome of patients medicated with 10–20 mg of finerenone (compared to eplerenone, using a composite endpoint of “death from any cause, cardiovascular hospitalizations, or emergency presentation for worsening HF” within 90 days) [51].

SGLT2 Inhibitors

Clinical trials investigating the therapeutic implications of SGLT2 inhibitors in HFpEF have focused primarily on the effects of dapagliflozin (NCT03619213, NCT03877224) and empagliflozin (NCT03057951, NCT03448406, IRCT20190122042450N2).

The EMPERIAL-preserved trial (NCT03448406) found that empagliflozin had no significant effects in patients’ exercise ability (measured through the 6MWTD), although treated patients displayed improvements in quality of life (measured through the KCCQ score), compared with placebo arm [52].The EMPEROR-preserved trial studied the effects of empagliflozin in a composite primary endpoint of cardiovascular death or hospitalization for HF in patients with HFpEF. This trial found that empagliflozin significantly reduced the risk of the primary endpoint in patients with HFpEF (hazard ratio 0.79, 95% CI (0.69–0.90)), regardless of the presence or absence of diabetes or patients’ EF [53]. Furthermore, it also showed that empagliflozin reduced the relative risk of first and recurrent hospitalizations for HF and significantly slowed kidney function decline. Thus, the EMPEROR-preserved trial has established empagliflozin as the first and only therapy, to date, to significantly reduce the risk of the composite of cardiovascular death or hospitalization for HF in adults with HFpEF. Nevertheless, it is important to note that empagliflozin’s effect seems to diminish at LVEF ≥ 60% (hazard ratio 0.87, 95% CI (0.69–1.10)), suggesting that it is ineffective for patients in the upper range of EF.

A recent clinical trial is currently evaluating the combination of dapagliflozin and low dose of pioglitazone on hospitalization rate and all-cause mortality in patients with HFpEF (NCT03794518). Although not yet confirmed for HFpEF, dapagliflozin has been shown to reduce the risk of HF hospitalizations and cardiovascular death in patients with HFrEF [54]. Furthermore, pioglitazone has been associated with lower risk of recurrent major adverse cardiovascular events, stroke, or myocardial infarction, even though it has been shown it does not reduce the risk for all-cause mortality and might even increase the risk of development of HF [55]; the combination of both these drugs could yield interesting results in HFpEF.

Phosphodiesterase Inhibitors

Phosphodiesterase 5A has been found to reverse cardiac remodeling in hearts subjected to sustained pressure load [56] and to improve contractile function, quality of life, and exercise capacity in small scale, randomized, double-blinded, placebo-controlled trials in patients with HFrEF [57-60], hinting towards a potential beneficial effect in patients with HFpEF. Different phosphodiesterase 5A inhibitors have been investigated in HFpEF: sildenafil (NCT01726049, NCT00763867), udenafil (NCT01599117), and tadalafil (DRKS00014595). Sildenafil has consistently failed to show beneficial effects in HFpEF. The RELAX trial (NCT00763867) found that phosphodiesterase 5A inhibition had no effect on maximal or submaximal exercise capacity, clinical status, quality of live, LV remodeling, diastolic function parameters, or pulmonary artery systolic pressure while also showing that treatment resulted in further worsening of patients’ renal function and led to increased levels of both NT-proBNP and uric acid [61]. A subsequent trial (NCT01726049) also reported no effects on hemodynamic parameters, such as mPAP, pulmonary capillary wedge pressure (PCWP), and cardiac output (CO), as well as no improvement in cardiac structure or function, cardiopulmonary exercise testing, laboratory parameters, or quality of life in patients with HFpEF and group 2 PH [62, 63]. Results regarding the trials with udenafil and enapril have yet to be reported.

Recently, a type III phosphodiesterase inhibitor, milrinone, has been evaluated for its hemodynamic effects in patients with HFpEF [64, 65]. Although milrinone showed no improvement on patients’ rate of isovolumic relaxation, LV stiffness, and minimal effect in end-diastolic pressure–volume relationships, it decreased right atrium pressure, mPAP, and PCWP during exercise suggesting that it might represent a relevant therapeutic option for HFpEF; however, pharmacological modulation of other cardiovascular parameters might be required to achieve optimal effects [64].

Prostaglandin Analogs

Prostaglandin analogs have been approved for the treatment of pulmonary arterial hypertension due to their vasodilatory effect [66]. Since PH due to left heart disease, and mainly HFpEF, is the most frequent cause of PH worldwide, prostaglandin analogs such as treprostinil have been evaluated for their effectiveness in subjects with PH associated with HFpEF (NCT03037580, NCT03043651). These trials were terminated by the sponsor due to slow enrolment, and due to the reduced number of subjects, efficacy-related endpoints were not analyzed, so its value as a novel therapeutic option for HFpEF remains unknown.

GLP-1 Analogs and GLP-Receptor Agonists

Small pilot studies in diabetic patients with HF (EF < 35%, NYHA III–IV) have found that GLP-1 analogs, such as exenatide, significantly increase patients cardiac index while decreasing PCWP shortly after infusion [67]. Continuous, 5-week infusion of recombinant GLP-1 was also associated with improved EF, Minnesota Quality of Life score, 6MWTD, and exercise peak VO2, effects similar in magnitude in both diabetic and non-diabetic patients [68]. A subsequent study found that GLP-receptor agonist, albiglutide, administration in subjects with EF < 40%, NYHA II–III, significantly improved peak VO2, but showed no effects in left ventricle (LV) size or function, 6MWTD, or quality of life scores [69]. Larger clinical trials such as the LIVE trial (NCT01472640) and the FIGHT trial (NCT01800968) have evaluated the effect of liraglutide, a GLP-receptor agonist, in patients with HFrEF. The LIVE trial found that liraglutide did not significantly affect patients’ systolic function but did result in weight loss, improved glycemic control, and improved physical performance [70]. It is important to note that serious adverse cardiac events occurred more often with liraglutide than with placebo [71]. The results from the FIGHT trial were neutral overall, showing no differences in outcomes, functional capacity, or post-hospitalization stability. Overall, these findings suggest that GLP-1 analogs and GLP-receptor agonist could show promising results in patients with HFpEF. Currently, three studies are evaluating both semaglutide’s and tirzepatide’s effects in patients with HFpEF and obesity and/or type 2 diabetes mellitus (NCT04788511, NCT04916470, NCT04847557).

Iron Products

Iron deficiency is a widespread comorbidity among HF patients [72], associated with longer hospital stays and higher healthcare costs [73]. While it has been thoroughly studied in patients with HFrEF, with strong evidence suggesting its association with decreased exercise capacity and quality of life, and treatment has both been tested and approved with demonstrated clinical benefit [74-77], there is less evidence when it comes to its association with HFpEF [78] with some studies suggesting it might be associated with reduced functional capacity and decreased quality of life [79-81]. The PREFER-HF trial (NCT03833336) evaluated the effects of iron therapy in patients with HFpEF and iron deficiency, although, to date, no results have been reported.

Anti-inflammatory Drugs

Because activation of inflammatory pathways has long been suggested to contribute to the pathogenesis of HF [82-84], some clinical trials have evaluated the effects anti-inflammatory drugs in patients with HF. A recent clinical trial has studied the efficacy of colchicine in patients with stable chronic HF (EF ≤ 40%) [85]. In this study, while colchicine was proven to be effective in reducing inflammatory biomarker levels, it did not affect patients’ functional status, regarding NYHA functional class or exercise tolerance. These results warrant attention to the newly initiated COLpEF (NCT04857931) trial, investigating colchicine in HFpEF, especially since the study’s primary outcome measures are changes in C-reactive protein, with no particular focus on improvement of patients’ cardiac functional status and symptoms.

β2 Adrenergic Receptor Agonists

Because pulmonary vascular resistance fails to decrease appropriately during exercise in patients with HFpEF, Reddy et al. hypothesized that drugs that enhanced pulmonary vasodilation, such as albuterol, could display a beneficial effect in these subjects (NCT02885636) [86]. In this trial, inhaled albuterol showed favorable effects on pulmonary vascular load during exercise, coupled with improvements in cardiac output reserve, right ventricular-pulmonary artery coupling, and left heart filling while maintaining pulmonary capillary hydrostatic pressures. Even though this study did not report LV functional responses to albuterol nor chronic effects, it suggests, overall, a possible role of β2 adrenergic receptor agonists in the treatment of HFpEF.

Hyperpolarization-Activated Cyclic Nucleotide-Gated Channel Blockers

There is one active trial on the effects of ivabradine in HFpEF (jRCTs051200059), for which there are still no published results. Going back to 2017, the EDIFY clinical trial (EudraCT no. 2012 002,742 20) showed that ivabradine-induced heart rate reduction failed to improve the following outcomes in HFpEF patients: echo-Doppler E/e′ ratio, 6MWTD, and plasma NT-proBNP concentration. Despite the disappointing results, this trial showed no concerns regarding the safety of the drug [87]. However, regarding the safety of ivabradine, in patients with coronary artery disease (but without HF), there was a 20% increase in HF-related hospital admissions [88]. As these diseases often come hand in hand, these results could be a cause for concern with the use of ivabradine in HFpEF.

Guanylate Cyclase Stimulators

Guanylate cyclase (GC) triggering by nitric oxide (NO) promotes vasodilation and inhibits smooth muscle cell proliferation, platelet aggregation, and vascular remodeling [89]. Since several cardiovascular diseases are associated with NO/GC-signaling pathway dysfunction [90, 91], GC stimulation could show potential benefits through the enhancement of the affinity of GC even at very low levels of NO [92].

Currently, 3 different GC stimulators are being studied in HFpEF: IW-1973 (NCT03254485), riociguat (NCT02744339), and vericiguat (NCT03547583 and NCT01951638). Of those, the VITALITY-HFpEF trial (NCT03547583) found that 24-week treatment with vericiguat at either 15 or 10 mg/day did not improve either the KCCQ physical limitation score or the 6MWTD, when compared with placebo [93]. These results contrast with those previously reported in the SOCRATES-PRESERVED trial (NCT01951638), where vericiguat, even with a smaller dosage than the one used in the VITALITY-HFpEF trial, was shown to improve patients’ KCCQ physical limitation score [94]. The differing results between these trials warrants attention and mandates further investigation.

NO-Donating Drugs

All past and ongoing trials using nitrates and nitrites are currently in phase II at most, having yet to show enough safety and effectiveness to warrant phase III trials to begin.

Despite the disappointing results of organic nitrates, inorganic formulations given in the form of nitrate-rich beetroot juice (12.9 mmol of NO3− in 140 mL) have been investigated for its effects in exercise capacity in patients with HFpEF (NCT01919177) [95]. It was found that patients receiving inorganic nitrate showed no changes in exercise efficiency (total work/total oxygen consumed), the trial’s primary endpoint. However, a single dose of inorganic nitrate prior to exercise significantly improved peak VO2 while also decreasing systemic vascular resistance and increasing CO at peak exercise. Overall, these results suggest some degree of improvement of exercise capacity in HFpEF patients with inorganic nitrate supplementation. Nevertheless, these should be confirmed in larger cohort studies that also evaluate inorganic nitrates’ long-term effects and its impact in parameters other that exercise capacity.

Inorganic nitrite has been recognized as an alternative source of NO-cGMP that is independent of the traditional NO synthase pathway [96-99]. Inorganic nitrite is reduced to NO particularly under conditions of tissue hypoxia and acidosis [98], suggesting it could selectively target hemodynamic alterations induced by stress in HFpEF [100, 101].

Several clinical trials have investigated the effects of inorganic nitrites in HFpEF. To our knowledge, the first study investigating inorganic nitrites in HFpEF (NCT01932606) found that intravenous sodium nitrite administration significantly improved exercise PCWP, resulting in a 37% reduction in left heart filling pressures with exercise [102]. Furthermore, nitrite therapy was associated with beneficial myocardial effects such as increased in LV stroke work with exercise, an integrated index of LV diastolic and systolic performance. Beneficial effects were of great magnitude during exercise compared with at rest. In another trial investigating nebulized inhaled sodium nitrite (NCT02262078), it was found that, similarly to the intravenous administration route, inorganic nitrate reduces PCWP both at rest and, particularly, during exercise [103]. However, a posterior trial (NCT02742129) found that inhaled sodium nitrate did not improve peak aerobic capacity, daily activity levels, or quality of life scores, contrasting with previous results and warranting attention to the drug as a HFpEF therapeutic option [104].

Several other clinical trials testing alternative formulations targeting the inorganic nitrate/nitrite pathway are currently under way—NCT02918552, NCT01919177, NCT03015402, NCT02980068, NCT02840799, NCT03289481, and NCT02713126.

Late Sodium Current Inhibitors

Since late sodium current is abnormally elevated in HF [105], and its inhibition improves diastolic performance in ischemic myocardium [106], there is ongoing effort to investigate the possible effects of ranolazine in HFpEF, with the RALI-DHF (NCT01163734) being the main trial for this research.

The RALI-DHF trial found that ranolazine improved hemodynamic measurements but had no effects in relaxation parameters [107]. It was found that ranazoline infused intravenously over 24 h resulted in immediate, albeit modest, improvements in left ventricle (LV) end-diastolic pressure, PCWP, and mPAP, suggesting a potential role in the treatment of diastolic dysfunction. Despite this, CO and stroke volume were decreased in the presence of ranazoline, pointing towards an acute reduction of systolic function, which could offset the positive effects of the drug on diastolic function. After 14 days of treatment, no significant changes were found in echocardiographic parameters or exercise tests, showing no evidence that acute changes induced by ranazoline would be predictive of long-term benefits.

Calcium Sensitizers

Cardiac troponin C acts as a Ca2+-operated molecular switch that turns myocardial force production on and off during systoles and diastoles [108]. Therefore, the kinetics and extent of contraction and relaxation of the heart are both coordinated by the Ca2+-binding characteristics of cardiac troponin C. Levosimendan is a Ca2+ sensitizer that, in patients with HFrEF, has been shown to produce dose-dependent increases of CO and decreases of PCWP, central venous pressure, peripheral vascular resistance, and systemic vascular resistance (NCT01536132, NCT00988806, NCT01065194) [109]. Because these effects would also be beneficial for patients with HFpEF, it has been recently evaluated in phase II trials (NCT03624010, NCT03541603).

The HELP trial (NCT03541603) has found that 24 h infusion of levosimendan in patients with PH in the setting of HFpEF resulted in significantly decreased PCWP and central venous pressure at rest, although these parameters were not altered during exercise [110]. Furthermore, submaximal exercise capacity, measured by 6MWTD, was also improved. These are encouraging findings that justify further study of the applicability of levosimendan in patients with PH in the setting of HFpEF.

Future Perspectives

Due to the complexity of the data and heterogeneity of patients, the identification of distinct clinical phenotypes using machine learning may allow for more targeted diagnostics and personalized therapeutic options [119]. Cohen et al. identified three distinct phenogroups that displayed differences in circulating biomarkers, cardiac/arterial characteristics, and prognosis among TOPCAT trial participants [120]. Interestingly, spironolactone therapy was associated with a more pronounced reduction in the risk of cardiovascular death, HF hospitalization, or aborted cardiac arrest in patients with more functional impairment, higher comorbidity burden, and the worse overall prognosis but did not appear to substantially benefit other phenogroups. In the absence of clear effective therapeutic options to improve prognosis and given the heterogeneity of risk factors and outcomes in HFpEF, the separation and identification of individuals into subgroups could aid the identification of patients who would mostly likely benefit from targeted interventions.

These nuances regarding the different subgroups of HF and the presence of different comorbidities could be the cause for some of the disappointing results in past clinical trials and need to be considered when designing future trials and tailoring future therapies.

The number of enrolled patients in some trials is often lackluster, creating the possibility that some beneficial therapies might go unnoticed because only of lack of statistical power. Not only this, but the endpoints of some of the trials need to be better defined, focusing more on clinical outcomes than on biochemical markers that in the end do not correlate as well as expected to the desired clinical outcomes. Furthermore, a confusing factor in the interpretation of these clinical trials is the heterogeneity in the LVEF thresholds adopted [121]. Current inclusion criteria range from ≥ 40% and > 40% to ≥ 45%, including patients with mildly reduced ejection fraction, considered by the European Society of Cardiology as heart failure with mid-range ejection fraction. The definition of LVEF threshold seems to be a relevant point because the largest benefits on the primary endpoints were recorded for LVEF ranging between 40 and 50%, while the same treatments were found to be ineffective for patients in the upper range of EF (> 60%) [121].

There is a need not only for new clinical trials results using different pharmacological classes, but also for more retrospective studies on the drugs currently empirically used for HFpEF without strong evidence, such as beta-blockers, on one hand to ensure patients are taking only the necessary drugs (as all have potential side effects) and on the other end of the spectrum to ensure clinicians that these drugs do not have deleterious cardiovascular effects when used for HFpEF.