LVDD is well-known pathophysiology of HFpEF. LVDD is caused by myocardial structural changes, microvascular dysfunction, systematic inflammation, and increased passive chamber stiffness [25]. The increased LVFP, the fundamental and pathologic hemodynamics of HFpEF, predominantly results from LVDD. The passive chamber stiffness is rapidly increased and the prolonged relaxation time is not shortened during exercise and these result in an increase of LVFP in HFpEF [26]. In the earlier stage, LVFP is normal but markedly increases during exercise, while in an advanced stage it increases continuously even at rest [27]. Elevated LVFP leads to elevation of left atrial (LA) pressure (passive LA hypertension) and contributes to an increase in pulmonary capillary hydrostatic pressure and an increase in vascular permeability, resulting in the development of interstitial edema [28]. Increased LVFP is closely related to the development of typical symptoms of HFpEF such as exercise intolerance and exertional dyspnea, symptom severity, and prognosis of HFpEF [29,30].

Under the condition of prolonged increased LVFP, secondary structural and functional remodeling of the LA develops. The preservation of LA function might be crucial to the adaptation of HFpEF, which prevents pulmonary congestion and right ventricular (RV) dysfunction [31]. Impaired LA function is related to exacerbation of pulmonary congestion, a change in lung function, and the development and worsening of pulmonary venous hypertension [31-33]. In a recent meta-analysis, it was proven that impaired LA function is valuable for diagnosis as well as for estimating the prognosis of HFpEF [34]. Especially in HFpEF patients with AF, LA structural and functional remodeling is exacerbated, which is a major contributor to disease progression, such as the development of pulmonary hypertension (PH), and RV dysfunction [35].

The presence of PH in HFpEF patients is associated with poor prognosis [21,36]. PH not only develops from passive LA hypertension but also from increased pulmonary vascular resistance, which is derived from pulmonary vascular disorders such as pulmonary vasoconstriction and pulmonary vascular remodeling [36-38]. A considerable number of HFpEF patients had pulmonary vascular disorders and increased mortality compared to patients with pure passive PH [36]. Regardless of the type of PH, persisting PH eventually induces RV dysfunction. The development of RV dysfunction is associated with a markedly increased risk of mortality [39].

The extrinsic restraint on the heart by pericardium and epicardial fat could lead to an increase in LVFP. Epicardial fat, as an external restraint, induces an increase in intracavitary pressure [40] that could promote inflammation, fibrosis, and coronary microvascular dysfunction (CMD) [20].

The left ventricular systolic dysfunction evaluated by tissue deformation imaging analysis is present at rest and worsens during exercise in HFpEF [41-43]. Abnormal ventricular-vascular coupling by arterial stiffening is regarded as the cause of subtle left ventricular systolic dysfunction. It results in a significantly impaired proportional increase in stroke volume output in response to exercise, which is one of the major causes of exercise intolerance in HFpEF. The increased left ventricular strain is closely associated with common comorbidities of HFpEF [44], and it has a prognostic value for CV mortality and HF hospitalization in HFpEF [42].

A considerable portion of patients presenting with symptoms of ischemia referred for coronary angiography are found to have evidence of ischemia but no significant stenosis of the coronary artery [45]. Most of these patients who have ischemia and no obstructive coronary artery disease (INOCA) have CMD [45]. In epidemiologic studies, INOCA was more prevalent in women and in patients with multiple comorbidities, which are similar to HFpEF [46]. Furthermore, patients with INOCA are likely to have preserved LVEF and a higher incidence of HF hospitalization [47]. Therefore, a link between CMD and HFpEF was strongly suggested [46], and the interaction between LVDD and CMD was clearly associated with an increased risk of HFpEF [48- 50]. CMD has become one of the major phenotypes and pathogenic cause of HFpEF.

Many patients with HFpEF display impaired cardiac output reserve during exercise, despite normal values at rest [43]. Not only left ventricular systolic dysfunction, as mentioned above, but also chronotropic incompetence results in a reduced cardiac output reserve [43].

Recent studies have identified an important role of the skeletal muscle in HFpEF. In patients with HFpEF, skeletal muscle mass is decreased [51], the composition of skeletal muscle changes with increased fatty infiltration [52] and the capillary density within the muscle is reduced [53]. The oxidative metabolism of skeletal muscle fiber changes to a slow type and a decreased mitochondrial content and abnormal fusion of mitochondria in skeletal muscle are observed [54]. These changes reduce peak O₂ uptake and result in exercise intolerance in patients with HFpEF.

Inappropriate activation of the renin-angiotensin-aldosterone system (RAAS) occurs in diseases that are associated with the development and progression of HFpEF, in common with HFrEF [55-58]. RAAS promotes the increase of arterial stiffness and myocardial stiffness and causes LVDD and LVH in HFpEF [55]. Therefore, several randomized clinical trials have been conducted to evaluate the prognostic value of RAAS blockade.

The angiotensin-converting enzyme inhibitor or angiotensin receptor blocker was not efficient for CV mortality and HF hospitalization in three large trials (CHARM-preserved [Candesartan in Heart Failure: Assessment of Reduction in Mortality and Morbidity], PEP-CHF [Perindopril in Elderly people with Chronic Heart Failure], and I-PRESERVED [Irbesartan in Heart Failure With Preserved Ejection Fraction]) [59-61]. They did not improve CV mortality, and only candesartan reduced HF hospitalization slightly [59]. A detailed summary of these trials is described in Table 4.

The Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist Trial (TOPCAT) is a large RCT to investigate the efficacy of aldosterone antagonists in symptomatic patients with HFpEF (LVEF ≥ 45%) [62]. The differences in the primary endpoint between spironolactone and placebo failed to reach a statistical significance; however, HF hospitalization was modestly decreased by spironolactone. Interestingly, in the subgroup analysis, spironolactone reduced primary outcome in patients with elevated natriuretic peptide levels at enrollment [62] in the Americas (hazard ratio [HR], 0.82; 95% CI, 0.69 to 0.98; p = 0.026) [63]. In Russia and Georgia, patients were enrolled by clinical judgment rather than having increased natriuretic peptides, and were relatively healthy and had lower compliance with study medication than in the Americas [64]. The result obtained in Russia/Georgia would not reflect a true response to spironolactone. For this reason, re-evaluation of the clinical efficacy of spironolactone in HFpEF is necessary. There is currently an ongoing trial in progress: Spironolactone Initiation Registry Randomized Interventional Trial in Heart Failure with Preserved Ejection Fraction (SPIRRIT, NCT 02901184), and their results are expected.

The angiotensin receptor-neprilysin inhibitor (ARNI) sacubitril-valsartan is a combination of inhibitors of neurohormonal activation and up-regulation of the adaptive natriuretic peptide pathway. In a phase-II study, (PARAMOUNT [65]) sacubitril-valsartan induced a greater decrease in N-terminal pro-B type natriuretic peptide (NT-proBNP), a larger reduction in LA size, and a greater improvement of symptoms than valsartan in patients with HFpEF. Therefore, the outcome trial, Efficacy and Safety of LCZ696 Compared to Valsartan, on Morbidity and Mortality in Heart Failure Patients With Preserved Ejection Fraction (PARAGON-HF) trial [66] was conducted in symptomatic HFpEF patients with increased natriuretic peptides. Despite a numerically lower event rate, the efficacy for HF hospitalization and CV death by sacubitril-valsartan approached but did not achieve a statistical significance (HR, 0.87; 95% CI, 0.75 to 1.01; p = 0.06). In post hoc analysis [13], the absolute risk reduction of sacubitril-valsartan was greatest in patients who were recently hospitalized within 1 month (approximately 25% to 30% risk reduction) and it gradually decreased with an increased interval from hospitalization. The sacubitril-valsartan might have alleviated the remaining neurohormonal activation after discharge. These data could provide clues for the initiation or switching time to sacubitril-valsartan in patients with HFpEF. The pre-specified analysis of outcomes by gender in the PARAGON-HF trial reported that the beneficial effect of sacubitril-valsartan was greater in women than in men (rate ratio [RR], 0.73, [95% CI, 0.59 to 0.90] in women vs. RR, 1.03, [95% CI, 0.84 to 1.25] in men, p interaction = 0.017) [67]. The possible reasons were further myocardial remodeling even in the same LVEF, more prominent age-related arterial stiffening in female patients with HFpEF, and differences in the signaling of natriuretic peptide [67]. Recently, Solomon et al. [68] reported the results of a pooled analysis of combined data from the Prospective Comparison of ARNI With ACEI to Determine Impact on Global Mortality and Morbidity in Heart Failure (PARADIGM-HF) [69] and PARAGON-HF [66] trials. The overall treatment benefit was at LVEF ≤ 42.5% and was maximized at lower ejection fraction. The sacubitril-valsartan was valuable in women with LVEF ≤ 60%, contrarily the threshold of LVEF at which efficacy of sacubitril-valsartan was highest was 45% to 50% in men. Therefore, it could be presumed that sacubitril-valsartan is effective for all patients with middle-ranged ejection fraction (HFmrEF). This result was consistent with the clinical characteristics of HFrEF and HFmrEF, which were similar and different from those of HFpEF [70].

There should be a careful application of RAAS blockade to patients with HFpEF because the phenotypes of patients in the real world are different from the inclusion criteria of trials. As the LVEF cutoff of trials varied, a considerable portion of registered patients to these trials might belong to HFmrEF. Neurohormonal activation is less prominent in HFpEF, rather than in HFmrEF or HFrEF. Furthermore, a large proportion of trial patients had already taken RAAS blocker (20% to 86% of the study population) and beta-blocker (55% to 80% of the study population) at enrollment [71], because of their comorbidities including CAD and arrhythmia and their neurohormonal system might be stabilized. Thus, it might leave a little room for the additional benefit from another RAAS inhibitor. Therefore, the messages of clinical trials should be interpreted and accepted judiciously.

Beta-blockers are the cornerstone of the management of HFrEF [8] because of a significant improvement in mortality and morbidity [72]. Sympathetic activation induces an increment of HR and leads to a shortening of the left ventricular diastolic filling time. Therefore, slowing HR and reversal of sympathetic overactivation could provide a benefit in HFpEF. A lower heart rate at discharge was closely related to a survival benefit in patients with HFpEF [73]. Several observational studies reported a modest benefit of beta-blockers for survival in HFpEF [74-76]. However, in pre-specified sub-analysis of the Study of Effects of Nebivolol Intervention on Outcomes and Rehospitalization in Seniors With Heart Failure (SENIORS) [77] the effect of beta-blockers nebivolol was similar between patients with HFrEF (LVEF ≤ 35%) and patients with HFpEF (LVEF > 35%), although the efficacy of nebivolol to reduce mortality was not statistically significant in both the patient groups. In the Japanese Diastolic Heart Failure Study (J-DHF) [78], an RCT to evaluate the effect of carvedilol in HFpEF (LVEF > 40%), it did not decrease CV death and HF hospitalization. Even in several meta-analyses, the efficacy of beta-blockers for mortality as well as HF hospitalization, in HFpEF is still controversial [79-81]. Recently, a pre-specified TOPCAT trial reported that the use of beta-blockers was related to an increase in HF hospitalization and was not associated with CV mortality in HFpEF [82]. The explained mechanism of ineffectiveness, even harm, was due to an increase in central blood pressure (BP) and myocardial wall stress due to increment of left ventricular volume and pressure by prolonged diastolic filling time [83,84]. However, recently reported Korean registry data showed that the use of beta-blockers reduced all-cause mortality in patients with HFpEF (LVEF ≥ 40%) and acute exacerbation of HF [85]. This benefit existed in patients with global longitudinal strain (GLS) < 14%, but not in patients with GLS ≥ 14% [86]. The conflicting results regarding the efficacy of beta-blockers are still unknown. This is due to the difference in the definition of HFpEF, especially LVEF cut-off value, heart rate at enrollment, and their changes after medication and pre-existing comorbidities, such as AF and CAD [87]. Despite these, beta-blockers might be beneficial to patients with HFpEF, especially in patients with myocardial remodeling with reduced GLS.

In altered endothelium-cardiomyocyte signaling by inflammation, NO is absent in cardiomyocytes. Therefore, organic NO supplementation was supposed to be helpful for HFpEF due to the restoration of myocardial NO content as well as alleviation of increased arterial load [5]. In the Nitrate’s Effect on Activity Tolerance in Heart Failure with Preserved Ejection Fraction (NEAT-HFpEF) trial [88], organic nitrate isosorbide mononitrate did not improve QOL and NT-proBNP, even if the daily activity level was reduced and HF symptoms worsened in the isosorbide mononitrate group. The possible causes were excessive hypotension and decreased cardiac output due to preload reduction [89], the rapid development of tolerance, and endothelial dysfunction by organic nitrates [90]. In contrast, inorganic nitrate (NO3) has a different NO metabolism: the nitrate-nitrite pathway. The nitrate-nitrite pathway could be an important route to restore NO in HFpEF, especially in the presence of hypoxia and acidosis such as skeletal muscle during exercise [91]. In small trials, the delivery of inorganic nitration via NO₃^--rich beetroot juice improved the exercise capacity of HFpEF patients [92,93]. The administration of sodium nitrite via infusion or inhalation improved cardiac output reserve, ventricular filling pressure, and pulmonary artery pressure during exercise in HFpEF. However, Inorganic Nitrite Delivery to Improve Exercise Capacity in HFpEF (INDIE-HFpEF) trial [94] showed that inhaled inorganic nitrite for 4 weeks failed to improve exercise capacity and QOL. The nitrite inhalation with its short-acting nature could be the cause of this result, as other administrations of inorganic nitrite, which provide a persistent and higher level of NO, achieved positive results. It is necessary to confirm the effectiveness of NO₃ in larger and long-term trials.

The sGC stimulators (riociguat and vericiguat) enhance cGMP production by acting on the NO receptor of sGC and activating sGC to generate cGMP. In Acute Hemodynamic Effects of Riociguat in Patients with Pulmonary Hypertension Associated with Diastolic HF (DILATE-1) [95] and Soluble guanylate Cyclase stimulatoR in heArT failurE patients with PRESERVED EF (SOCRATES-PRESERVED) [96] trials, sGC stimulators showed limited improvement in hemodynamic and echocardiographic parameters.

Despite the possible positive effect of sildenafil in HFrEF [97], two separate studies for evaluating the efficacy of sildenafil in HFpEF with PH or RV dysfunction failed to show improvement in exercise capacity and QOL, and sildenafil was associated with impairment of renal function and increment of neurohormone level [98,99].

PKG is an intrinsic suppressor of ventricular hypertrophy and interstitial fibrosis [5]; therefore, stimulation of PKG is a potential therapy in HFpEF. cGMP is mandatory to stimulate PKG. As mentioned above, cGMP is synthesized by sGC via the NO-sGC-cGMP-PKG pathway and is generated by receptor guanyl cyclase linked to natriuretic peptide receptors. Therefore, the sacubitril, a neprilysin inhibitor, stimulates PKG by elevating the level of natriuretic peptides. The efficacy of ARNI in HFpEF is described above.

Chronotropic incompetence contributes to impaired cardiac output reserve, as described above. The restored normal heart rate response during exercise by pacemaker might be beneficial to HFpEF with chronotropic incompetence. The Rate-Adaptive Atrial Pacing In Diastolic Heart Failure (RAPID-HF) trial is ongoing to evaluate the impact of rate-adaptive atrial pacing on exercise capacity (NCT 02145351).

Hypertension is the most common comorbidity of HFpEF patients. In many previous RCTs of antihypertensive medication, optimal treatment of hypertension reduced the incidence of HF [106]. In HFpEF, the additional benefit of lowering BP is uncertain. As mentioned above, there was discordance between BP lowering and clinical outcome in large trials with neurohormonal inhibitors in HFpEF, even though neurohormonal inhibition had a favorable effect on ventricular hypertrophy, interstitial fibrosis, and myocardial stiffness [5]. Nevertheless, a recent meta-analysis showed that BP lowering was closely related to the reduction in HF hospitalization rather than all-cause mortality in HFpEF patients [107]. Furthermore, optimal BP control is important to prevent other major CV outcomes, including stroke and CAD, even in HFpEF patients.

The diabetes is prevalent in 20% to 40% of HFpEF patients [21]. Three RCTs on sodium-glucose cotransporter-2 (SGLT2) inhibitors in type 2 diabetes with high CV risk or established CV disease demonstrated a valuable effect of SGLT2 inhibitors [108]. McMurray et al. [109] reported that dapagliflozin significantly reduced worsening HF or CV death in HFrEF patients. The benefit of dapagliflozin on HF was similar between the patients with and without diabetes. The potential mechanism of the beneficial effect of SGLT2 inhibitors is not only to decrease intravascular volume via osmotic diuresis and natriuresis but also to reduce neurohormonal activation [110]. SGLT2 inhibitors increase metabolic efficiency and myocardial energy supply [110]. Furthermore, SGLT2 inhibitors induce a decrease in oxidative stress and fibrosis, an increase of endothelial function, and vascular compliance, which may be favorable in HFpEF [110,111]. The diastolic parameter was improved by the administration of SGLT2 inhibitors in patients with type 2 diabetes patients [112,113]. The RCTs for the efficacy of SGLT2 inhibitor in HFpEF (EMPagliflozin outcomE tRial in Patients With chrOnic heaRt Failure With Preserved Ejection Fraction [EMPEROR-Preserved; NCT 03057951], Effect of EMpaglifozin on ExeRcise ability and heart failure symptoms, In patients with chronic heArt faiLure with preserved ejection fraction [EMPERIAL-preserved; NCT 03448406], Dapagliflozin Evaluation to Improve the LIVEs of Patients With PReserved Ejection Fraction Heart Failure [DELIVER; NCT 03619213], dapagliflozin in PRESERVED Ejection Fraction Heart Failure [PRESERVED-HF; NCT 03030235]) are ongoing and we look forward to the results of these studies shortly.

Lifestyle modifications, including exercise, weight reduction, and dietary composition, are beneficial in preventing and treating HF [114]. However, the potential benefit of lifestyle modification has not been fully elucidated in RCT or cohort studies. Nevertheless, the efficacy of exercise was reported in several studies that included endurance and resistance training, improved exercise capacity, and QOL in HFpEF [115,116]. About half of HFpEF patients have obesity, and an increase in body adiposity triggers systemic inflammation and impairment of cardiac, vascular, and skeletal muscle function [5]. Kitzman et al. [117] demonstrated that caloric reduction during 20 weeks significantly improved peak O₂ consumption, symptoms, and QOL in older and obese patients. The additive benefit was derived by a combination of caloric restriction and exercise. Lifestyle modification might be important for the management of HFpEF as well as a cardiometabolic syndrome, representative comorbidity of HFpEF. Further investigation to prove the beneficial effect of lifestyle modification in HFpEF is required.