Inflammatory Mechanisms in Heart Failure with Preserved Ejection Fraction

Abstract

Heart failure with preserved ejection fraction (HFpEF) is now the most common form of heart failure and a significant public health concern for which limited effective therapies exist. Inflammation triggered by comorbidity burden is a critical element of HFpEF pathophysiology. Here, we discuss evidence for comorbidity-driven systemic and myocardial inflammation and the mechanistic role of inflammation in pathological myocardial remodeling in HFpEF.

Introduction

Cardiovascular diseases remain a leading cause of death globally marked by an ever-evolving landscape of challenges (1). As we have become increasingly successful in treating acute cardiovascular events, attention has turned to heart failure (HF) as the next major hurdle in cardiovascular medicine; more and more people are surviving the acutely lethal manifestations of cardiovascular diseases, returning to family and society with an injured heart, often culminating in HF.

The syndrome of HF can be broadly categorized into two clinically distinct subtypes: heart failure with reduced ejection fraction (HFrEF) and heart failure with preserved ejection fraction (HFpEF). A growing therapeutic arsenal has improved outcomes in HFrEF, yet our ability to treat HFpEF remains modestly effective. In light of the significant morbidity/mortality associated with HFpEF, coupled with the fact that it is now the most prevalent form of HF, deciphering the molecular underpinnings of this syndrome is paramount (2–4). Leveraging these findings to promote the development of new therapies is critical for improving patient outcomes.

HFpEF is characterized by multiple phenotypic disorders, including diastolic dysfunction mediated by passive myocardial stiffness, microvascular dysfunction, metabolic dyscrasias, and more (5). With respect to myocardial stiffness, underlying cellular and molecular changes remain poorly understood; however, a critical role for inflammation is emerging. Initial observations correlating elevated serum proinflammatory cytokine levels and adverse clinical outcomes in both HFpEF and HFrEF sparked interest in inflammation as a driver of HF (6–9). It is now believed that this inflammation is largely generated by comorbidity burden and plays a critical role in HFpEF myocardial remodeling (3, 4). Indeed, the majority of HFpEF patients are obese and hypertensive, with advanced age, chronic kidney disease, and chronic obstructive pulmonary disease also being highly prevalent (5, 10).

In this review, we discuss mechanisms of comorbidity-driven inflammation and its importance in HFpEF pathogenesis. We delve into how inflammation promotes the characteristic cellular and molecular changes seen in HFpEF.

HFpEF in Perspective: from Disease to Heterogenous Syndrome

Early simplistic views of HFpEF as a disease resulting from abnormalities in diastolic function have morphed into a more complex understanding that HFpEF is a comorbidity-driven clinical syndrome. Each comorbidity contributes uniquely to overall syndrome pathophysiology. In fact, patients diagnosed with HFpEF are a heterogenous group presenting with varying clinical manifestations, serum markers, and comorbidity burden. Furthermore, HFpEF not only involves the heart and cardiomyocytes but is a systemic syndrome affecting multiple organs and cell types via diverse pathophysiological processes.

In an attempt to subclassify this heterogenous population, it has become increasingly popular to classify HFpEF patients based on comorbidity phenogroups, which include autoimmune/inflammatory disease-associated HFpEF, cardiorenal HFpEF, cardiometabolic HFpEF, likely to be the most prevalent form, and several more (4). Evidence for the existence of these subgroups derives largely from pheno-mapping studies based on machine learning-derived clustering analysis (11–15). The importance of characterizing these phenogroups lies in their differing prognoses and responses to therapy (15, 16). Using a precision medicine-based approach to treat individual subgroups may provide better outcomes in future clinical trials.

Inflammation at the Heart of HFpEF Pathophysiology

Levine et al. (17) first linked inflammation to HF, reporting an elevation in circulating tumor necrosis factor (TNF) in patients with HFrEF. This work was corroborated over the following decades by multiple studies showing a strong correlation between various serum proinflammatory cytokines and adverse clinical outcomes in HF (7, 18–20). These findings ushered in strong interest in exploring inflammation as a key pathophysiological contributor to HF, culminating in a multitude of both successful and unsuccessful phase III clinical trials.

Inflammation in HFpEF is characterized as chronic and low grade, present at both the systemic and myocardial levels. Studies based on animal models and patient endomyocardial biopsies have led us to propose the following: initially, there is an increase in both circulating and myocardium-infiltrating immune cells, including macrophages (21). This increase in leukocyte migration may be mediated by increased adhesion receptor expression and activation (e.g., E-selectin and ICAM) (22). Next, evidence exists for a potential imbalance in inflammatory to anti-inflammatory immune cell populations such as circulating T cells (Th17/Treg ratio); however, more work is needed to confirm this imbalance (23). Increased inflammatory activity of immune cells results in an increase in systemic and myocardial proinflammatory cytokine abundance in HFpEF (23–25).

In its initial formulation, the comorbidity-inflammation paradigm stipulated that inflammation in HFpEF is largely comorbidity driven, resulting in endothelial inflammation and coronary microvascular dysfunction limiting nitric oxide (NO) bioavailability and PKG activity in cardiomyocytes. The resulting cardiomyocyte hypertrophy and reduced distensibility combined with ventricular fibrosis due to increased myofibroblast activity have been held to contribute to HFpEF pathogenesis (26). Whereas reduced endothelial NO production has yet to be definitively linked to reduced cardiomyocyte PKG activity in HFpEF, evidence continues to accumulate implicating comorbidity-driven inflammation in a wide range of the pathophysiological events observed in this syndrome. Recent proteomics mediation analysis from the PROMIS-HFpEF study further bolsters the comorbidity-inflammation paradigm by revealing that inflammation mediates the association between comorbidity burden and echocardiographic parameters in HFpEF (27). Furthermore, tumor necrosis factor receptor 1 (TNFR1), urokinase plasminogen activator receptor (UPAR), insulin-like growth factor binding protein 7 (IGFBP7), and growth differentiation factor-15 (GDF-15) have all been identified as strong mediators of this association (27) (FIGURE 1).

FIGURE 1.

Heart failure with preserved ejection fraction-associated comorbidities increase circulating and activated, myocardium-infiltrating leukocytes Adipose tissue, in turn, contributes to circulating inflammatory signals. The resulting systemic and myocardial inflammation is characteristic of heart failure with preserved ejection fraction (HFpEF). HSC, hematopoietic stem cell.

Other bioinformatics studies have shed light on the importance of inflammation in overall syndrome severity and clinical outcomes. Clustering analysis performed with inflammatory biomarkers as input variables identified three inflammatory clusters in HFpEF: pan-inflammatory phenotype, noninflammatory phenotype, and obese-high CRP phenotype (14). Interestingly, patients with the pan-inflammatory phenotype exhibited the worst outcomes (14). Findings such as these highlight the likely importance of inflammation in HFpEF pathophysiology and identify patient subgroups that may benefit from anti-inflammatory therapies.

Cellular and molecular differences between the pathophysiologies of HFpEF and HFrEF continue to emerge with recent studies indicating that inflammation may play a more crucial role in HFpEF. Using data from the BIOSTAT-CHF program, network analysis enriched with data regarding protein-protein interactions revealed that pathways relating to inflammation and extracellular matrix (ECM) remodeling are overrepresented in HFpEF compared to HFrEF (28). Evidence remains inconclusive, however, as another study using transcriptomics data revealed similar importance of fibrosis, inflammation, hypertrophy, and oxidative stress in HFpEF and HFrEF, whereas protein homeostasis, endoplasmic reticulum stress, and angiogenesis were identified as unique to HFpEF (29). Most striking, however, was a clear distinction in gene expression patterns between HFpEF and HFrEF and the identification of a HFpEF phenogroup with inflammatory and matrix signatures (29). It is important to note that some of these differences may be attributable to the use of serum biomarkers in the former study and endomyocardial biopsies in the latter.

Mechanisms of Comorbidity-Driven, Low-Grade, Chronic Inflammation

Adipose Tissue Expansion

The majority of HFpEF patients are obese with an average body mass index above 35 kg/m² (30). The roles of different adipose tissue (AT) depots in the pathogenesis of HFpEF are poorly understood, yet AT expansion and associated inflammation are recognized as important drivers of metabolic dysregulation and systemic inflammation. Visceral AT (VAT) expansion and inflammation occur simultaneously. Deciphering how these factors trigger this inflammatory transformation is incomplete but includes AT hypoxia-inducible factor 1α (HIF-1α) activation, accumulation of proinflammatory long chain fatty acids in adipocytes triggering various inflammatory pathways, and possibly metabolic endotoxemia (31–36). The consequence is a well-characterized accumulation and activation of diverse inflammatory cell types in VAT. The ensuing release of a multitude of inflammatory molecules translates into increased systemic and myocardial inflammation (31).

Whereas extensive evidence implicates visceral adipose depots in both systemic and myocardial inflammation in HFpEF, epicardial AT (EAT) has been less well studied. Notably, due to its anatomical proximity and intricate association with the myocardium, EAT expansion and associated inflammation have been proposed as important elements in HFpEF pathobiology (37). Clinically, it is observed that in patients with cardiometabolic HFpEF, EAT expansion is associated with impaired hemodynamic parameters and worse exercise capacity (38). In fact, EAT accumulation has been independently associated with all-cause mortality and HF hospitalizations in HFpEF patients (39).

The epicardial space is connected to the myocardium by a direct microcirculatory network allowing the transfer of adipokines and immune cells (40–42). EAT expansion occurs concurrently with adipocyte transformation to a predominantly white phenotype and the adoption of a proinflammatory state (40, 43). This transformation predisposes epicardial adipocytes to lipolysis and secretion of proinflammatory cytokines (including leptin, TNF-α, IL-1β, and IL-6) while decreasing the secretion of adiponectin (44–46). A direct relationship has been observed between epicardial adipose thickness and both myocardial and systemic inflammation in obese animals (47). EAT can also act as a depot of both innate and adaptive immune cell populations, including macrophages, mast cells, and B and T cells located in fat-associated lymphoid clusters (FALCs) and in the stromal vascular fraction of EAT (42, 48, 49). EAT-resident immune cells such as macrophages in HFpEF begin to display a proinflammatory phenotype with EAT expansion and inflammation (47).

Accumulating evidence points to obesity-induced metabolic stress as a driver of inflammation, a process termed “meta-inflammation.” Meta-inflammation involves bidirectional cross talk between inflammation and metabolic processes, whereby metabolic intermediates affect inflammatory pathways and vice versa (3, 4). Metabolic alterations observed in obesity can fuel immunologic activation via a wide range of metabolic intermediates that modulate multiple cellular pathways (50). This is especially evident in immune cell populations in which metabolism can robustly affect inflammatory activity (51). TCA cycle derangements, linked with obesity, can lead to the accumulation of various TCA cycle intermediates triggering macrophage proinflammatory polarization and activation of inflammatory pathways (52, 53). Succinate accumulation, for example, promotes macrophage polarization toward an M1-like phenotype and activates HIF-1α, upregulating IL-1β expression (54). Similarly, fumarate accumulation increases TNF-α and IL-6 expression by inducing epigenetic modifications through KDM5 histone demethylase inhibition (55). This bidirectional immune-metabolic cross talk also results in inflammation-induced metabolic alterations. Of note, whereas these have not been extensively explored in a preclinical model of HFpEF utilizing high-fat diet (HFD) and IL-6 infusion, AMP-activated protein kinase (AMPK) is suppressed, and glucose metabolism is impaired (56) (FIGURE 2).

FIGURE 2.

Depicted are common comorbidities associated with HFpEF Each comorbidity is associated with, and drives, low-grade, systemic inflammation.

Hemodynamic Overload

Hypertension is the most prevalent comorbidity in HFpEF, affecting the majority of patients with the syndrome (57–59). There is increasing evidence that hypertension-induced immune activation is a central element mediating myocardial damage in hypertension (60). As in HFpEF, hypertensive patients harbor increased circulating proinflammatory cytokines, including C-reactive protein (CRP), TNF-α, IL-1β, and IL-6 (61, 62).

Multiple animal models of hypertension manifest increased capillary permeability and immune cell recruitment to the myocardium with evidence of immune-mediated myocardial damage (63–65). Both innate and adaptive cellular immunity play crucial roles in mediating these effects. Like HFpEF, an imbalance in T-cell populations (Th17/Treg) skews the overall T-cell phenotype to proinflammatory. Dendritic cells play an important role in this imbalance by inducing polarization of T cells into proinflammatory Th17 cells, as was shown in the deoxycortisone acetate-salt (DOCA-salt) murine model (66). In fact, the presence of Th17 cells is necessary for the development of hypertension in angiotensin II (ANG II) models (67).

HTN can induce Nod-like receptor-containing pyrin domain 3 (NLRP-3) activation in antigen-presenting cells resulting in immune activation (68). Predominance of M1-like over M2-like macrophages in spontaneously hypertensive rats points to proinflammatory polarization of macrophages in hypertension (69, 70). Tissue infiltration of these inflammatory macrophages can then induce end-organ damage, increased CD11b⁺ macrophage adhesion, and migration in hypertension. Interestingly, the presence of these macrophages is also essential for development of hypertension (71).

Hypertension-induced myocyte damage promotes the release of damage-associated molecular patterns (DAMPs) capable of activating inflammatory cascades in both immune cells and in cardiomyocytes (72). DAMPs include high mobility group box 1 (HMGB1), heat shock proteins, DNA fragments, oxidized LDL, and mitochondrial content. DAMPs induce an inflammatory response via Toll-like receptors (TLRs), important mediators of inflammation in cardiovascular diseases (CVD) (73). In fact, hypertension results in increased cardiac TLR-4 expression and activity, contributing to chronic, myocardial, low-grade inflammation (74).

Aging

A strong association exists between aging and HFpEF, with the incidence of HFpEF doubling with every decade of life after 65 (75). Aging-associated systemic, chronic, low-grade inflammation, also known as “inflammaging,” has been recognized as an important pathophysiologic process increasing CVD risk in aging (76). Inflammaging is evidenced by the presence of elevated circulating proinflammatory cytokines, including IL-1, IL-6, IL-8, transforming growth factor-β (TGF-β), and TNF-α without the presence of active disease or comorbidities (77–80).

Secretion of senescence-associated secretory phenotype (SASP) signals by aging cells has emerged as a key driver of both local and systemic inflammation and a paracrine effector mediating senescent cardiomyocyte interaction with other myocardial cell populations. SASP signals include a wide range of bioactive molecules, such as proinflammatory cytokines, proteases, and extracellular matrix components that are secreted by senescent cells (81, 82). A specific proinflammatory subset of SASP signals that include chemotactic factors [monocyte chemoattractant protein 1–4 (MCP 1–4], proinflammatory cytokines (IL-1β, TNF-α, IL-6, etc.), DAMPs (e.g., HMGB1, DNA fragments, oxidized cardiolipin, etc.), and immune cell differentiation and maturation factors [macrophage colony-stimulating factor (M-CSF), granulocyte macrophage-colony-stimulating factor (GM-CSF), etc.] can contribute to inflammation (76). Activin, another proinflammatory SASP signal, for example, exacerbates age-related cardiac dysfunction (83). Interestingly, SASP-induced proliferation of CD38⁺ macrophages contributes to reductions in NAD⁺ levels seen in aging (84). As we have shown previously, NAD⁺ deficiency is an important feature of HFpEF (85).

Aging results in the infiltration of inflammatory cells into the myocardium, an indicator of myocardial inflammaging (21). Clonal hematopoiesis of indeterminate potential (CHIP) is another mechanism promoting immune dysregulation and subsequent cardiac dysfunction in aging (86, 87). Clinical studies have revealed a link between CHIP and increased 5-yr mortality in HF (88, 89). Numerous experimental studies have also linked CHIP driven by Dnmt3a, Tet2, or Jak2 to HF (87, 90–92), although a recent study has called into question CHIP-induced inflammation as a driver of CVD (93). Remarkably, Tet2 CHIP alone results in spontaneous cardiac dysfunction in aged mice (87, 92). This mutation has been directly linked to increased NLRP3/IL-1β signaling in the heart (91).

Chronic Kidney Disease

Like obesity and aging, chronic kidney disease (CKD) is associated with a persistent, systemic, proinflammatory state. As CKD progresses and the glomerular filtration rate (eGFR) declines, there is a proportional increase in inflammatory cytokines and acute phase reactant levels (94). Remarkably, in the background of CKD leading to increased levels of proinflammatory cytokines, acute phase reactants, or increased IL-1α expression in circulating monocytes, is associated with a higher risk of CVD (95, 96). Interestingly, the preclinical murine model of salty drinking water-unilateral nephrectomy-aldosterone (SAUNA) provides a clear indication that kidney-related disease can result in diastolic dysfunction, myocardial inflammation, and pathologic myocardial remodeling (21, 97–99).

Various factors have been implicated in driving systemic inflammation in CKD (100, 101). Increased production and decreased renal clearance of inflammatory cytokines contribute to elevated systemic levels (102). Uremia drives oxidative stress which in turn drives inflammation (103). Metabolic acidosis, a hallmark of CKD, and sodium overload can also induce inflammation (104, 105). Sympathetic nervous system stimulation (SNS) and renin-angiotensin-aldosterone (RAAS) activation, hallmarks of both HFpEF and CKD, have also been implicated in inflammation via multiple pathways (65, 106). Whereas both SNS and RAAS blockade strategies have failed in HFpEF, these two pathways remain contributors to overall inflammation.

Increased levels of circulating carbamylated LDL, a posttranslational modification of LDL that increases its inflammatory properties, have been reported in CKD (107, 108). Carbamylated LDL has been implicated in endothelial dysfunction and a subsequent decrease in endothelial nitric oxide (NO) release, a primary pathophysiologic mechanism in HFpEF (109). Similarly, symmetric dimethylarginine (SDMA) accumulates in HDL and has been implicated in TLR2-mediated endothelial inflammation and endothelial nitric oxide synthase (eNOS) inhibition (110–112).

Accumulation of proinflammatory microbiota-derived trimethylamine N-oxide (TMAO), an independent risk factor of HFpEF and predictor of renal dysfunction in HFpEF, occurs in CKD (113). A similar increase in other proinflammatory gut-derived toxins is also seen (114). SASP-driven inflammation occurs in CKD due to premature cellular senescence evidenced by the accumulation of cells expressing p16ink4a or p21cip1 (115).

Inflammation-Driven HFpEF Pathophysiology

Inflammation-Induced Extracellular Matrix Remodeling

Myocardial “reactive” fibrosis driven by ECM remodeling plays a key role in altering the mechanics of myocardial diastolic function (116). ECM remodeling is governed fundamentally by cardiac fibroblasts via the synthesis of fibrillar ECM proteins and secretion of degradative enzymes [matrix metalloproteases (MMPs) and ADAMTS] and regulators of collagen crosslinking (e.g., SPARC, osteopontin, lysyl oxidase) eliciting a net increase in ECM deposition (117).

Immune cell-fibroblast cross talk is a major driver of the differentiation of quiescent fibroblasts into activated myofibroblasts and thus a key element of pathological myocardial ECM remodeling in HFpEF. Various immune cell types, including macrophages, dendritic cells, mast cells, and B and T cells, have been implicated in fibroblast activation and myocardial fibrotic remodeling. An association between collagen deposition, immune cell infiltration, and diastolic dysfunction has been observed in HFpEF patients (118). In both HFpEF endomyocardial biopsies and different HFpEF preclinical models, it has been shown that both myocardial macrophage and neutrophil populations expand in HFpEF. Furthermore, macrophage-specific knockout of the profibrotic molecule IL-10 improved diastolic dysfunction and reduced fibrosis in these models (21). Inhibition of immune cell myocardial infiltration may provide a promising strategy to impede fibroblast activation and myocardial fibrosis. In fact, inhibition of macrophage migration to the myocardium with an anti-ICAM-1 monoclonal antibody (Nab) reduced myocardial fibrosis by inhibiting TGF-β expression and fibroblast proliferation (119). Similarly, treatment of ANG II-infused mice with MCP-1 monoclonal antibodies reduced myocardial macrophage accumulation and ameliorated diastolic dysfunction (120). As immune cells infiltrate the myocardium, the presence of a proinflammatory milieu triggers further inflammatory activation.

A variety of proinflammatory cytokines including IL-6, the IL-1 family, and IL-18 can induce fibroblast activation, several acting by converging on the TGF-β signaling pathway, a central mechanism governing fibroblast activation (121–123). TGF-β acting via both canonical and noncanonical pathways is central in downregulating MMPs, upregulating tissue inhibitors of matrix metalloproteases (TIMPs), and upregulating ECM protein expression, tipping the balance toward ECM expansion (124). Interestingly, elevations in the fibroblast activator short suppression of tumorigenicity 2 (sST2) and TIMP-1 have been associated with worse outcomes in HFpEF patients (125). Proinflammatory cytokines including TGF-β can also induce endothelial-to-mesenchymal transition (EndMT), which may be an important contributor to myocardial fibrosis in HFpEF (126).

Activation of cardiac fibroblasts can, in turn, contribute to myocardial inflammation provoking a positive feed-forward loop. NF-κB activation in myofibroblasts has been shown to increase Ly6C^hi monocyte recruitment to the myocardium (127). In fact, cardiac fibroblasts can function as antigen presenting to CD4⁺ T cells via IFN-γ induced major histocompatibility complex type II (MHC II) (128). This points to a bidirectional cross talk between fibroblasts and immune cells that amplifies inflammation and fibrosis.

Whereas the PIROUETTE trial testing the anti-TGF-β drug pirfenidone in HFpEF revealed reduced ECM volume, there was no change in diastolic function (129). Limited reduction in ECM volume may underlie a lack of improvement in diastolic dysfunction. Interestingly, pirfenidone has anti-inflammatory properties and may still emerge as a useful tool (130, 131). Whereas traditional methods of detecting fibrosis have revealed more modest degrees of fibrosis in HFpEF than previously thought, a potentially less recognized role of “hidden fibrosis” resulting from an increase in ECM collagen deposition detectable by highly sensitive methods, such as quantitative mass spectrometry, may play an important role in diastolic dysfunction (24, 132, 133). Highly sensitive methods of detection of myocardial fibrosis may help to further decipher ECM remodeling in HFpEF. Furthermore, alternative strategies targeting fibrosis may prove fruitful. Chimeric antigen receptor T cell (CAR-T) therapy has been used to transiently target fibroblasts, reduce fibrosis, and improve cardiac function in different HF models (134, 135). Similarly, cardiosphere-derived cells (CDCs) reduced inflammatory infiltrates, myocardial fibrosis, and ventricular arrhythmias in Dahl salt-sensitive rats (136, 137).

Groundbreaking results from the EMPEROR-Preserved trial indicate empagliflozin can reduce HF hospitalizations in HFpEF (58). Among the diverse effects of SGLT2 inhibition is the ability to attenuate both systemic and myocardial inflammation. Recently, increased expression of HMGB1, a DAMP, was reported to be critical to neutrophil extracellular trap (NET) formation in HFpEF. This HMGB-NET axis can mediate diastolic dysfunction and myocardial fibrosis and is inhibited with SGLT2 inhibitor therapy (138). Similarly, SGLT2 inhibition has been shown to reduce systemic inflammation in both animal models and patient populations with diabetes (139–142). Whereas anti-inflammatory mechanisms remain under investigation, SGLT2 inhibitors have been shown to reduce NLRP3 activation in various tissues including the heart (143–146). Empagliflozin-mediated suppression of increases in β-hydroxybutyrate can block NLRP3 activation in diabetic individuals at high risk of cardiovascular disease (147). A similar effect on NLRP3 has been reported in rodent models of HF (146). Empagliflozin has also been shown to promote adipose tissue browning and reduce inflammatory M1 macrophage accumulation in adipose tissue, possibly limiting metabolic inflammation (147). Myocardial antifibrotic effects have also been reported in several models of cardiovascular disease (148–150) (FIGURE 3).

FIGURE 3.

Inflammation underpins pathological cardiomyocyte and fibroblast alterations that contribute to myocardial stiffness in HFpEF NO, nitric oxide.

Inflammation-Induced Cardiomyocyte Dysfunction

Myocardial stiffness in HFpEF is multifactorial, and inflammation-driven alterations in cardiomyocyte morphology, homeostasis, distensibility, and electrophysiology are important contributors. Inflammation and nitro-oxidative stress are closely linked and important contributors to HFpEF pathobiology. NO signaling via iNOS is induced by inflammation and plays a fundamental role in mediating cardiomyocyte dysfunction in HFpEF. As we have reported in both patient samples and in a HFD plus N^ω-nitro-l-arginine methyl ester (l-NAME ) mouse model of HFpEF (151), upregulation of cardiomyocyte inducible nitric oxide synthase (iNOS) leads to S-nitrosylation of the endonuclease inositol-requiring protein 1α (IRE1α) resulting in defective XBP1 splicing, accumulation of unfolded proteins, and cardiomyocyte dysfunction (151, 152).

The mitochondrial respiratory chain is the primary source of oxidative stress in CVD, with other mitochondrial proteins, such as NADPH oxidases (NOX2/4), xanthine oxidase, and cytochrome P450 enzymes contributing to the reactive oxygen species (ROS) pool (153). Inflammation can drive ROS production and vice versa, triggering a vicious cycle of inflammation and oxidative stress (154). Proinflammatory cytokine signaling is a culprit in macrophage-mediated inflammation and has been implicated in HFpEF in driving mitochondrial ROS via IL-1β (155). TGF-β can increase NOX4 activity and reduce antioxidant factors such as glutathione, driving ROS accumulation (156). IL-6 and TNF-α have been similarly implicated in increasing oxidative stress (154, 157).

Titin compliance is important in cardiomyocyte distensibility. Coronary endothelial inflammation resulting in dysregulation of the eNOS-cGMP-PKG axis with consequent hypophosphorylation of cardiomyocyte titin is a proposed mechanism of reduced passive cardiomyocyte elasticity in HFpEF; however, it is important to note that cardiomyocytes have an innate ability to stimulate PKG via NO production. Activation of PKG or PKA in vivo improves left ventricular (LV) chamber compliance via titin N2B segment phosphorylation (158). Interestingly, activation of PKG has a wide range of effects on the myocardium, including reduced inflammation, hypertrophy, and fibrosis (159–161). In addition, PKG is known to stimulate autophagy, improve mitochondrial function, and increase the clearance of misfolded proteins via proteasomal degradation, making it an attractive therapeutic target (162–164). Several phosphodiesterase (PDE)-5 and -9 inhibitors that augment cardiomyocyte cGMP, and thereby PKG activity, have been examined in multiple preclinical models of HFpEF (165–168). Whereas clinical trials using PDE5 inhibitors have revealed no improvement in patient clinical status, we are still awaiting results from PDE9 inhibitor trials (169). Important differential actions of PDE5 and PDE9 exist due to effects on NO-stimulated cGMP in the former and natriuretic peptide-stimulated cGMP in the latter (167, 170). Alternatively, targeting inflammation may prove to be successful in attenuating endothelial dysfunction and dysregulation of the eNOS-cGMP-PKG axis. The IL-1 antagonist anakinra reduced endothelial NOX and NF-κB activation and subsequently endothelial dysfunction in streptozotocin-induced diabetic rats (171). Whereas cytoskeletal α-tubulin abundance/posttranslational modifications and titin isoform ratios have been linked to increased cardiomyocyte stiffness in HFpEF, no link has been made with inflammation (172–174).

Proinflammatory cytokine signaling can alter cardiac electrophysiology contributing to abnormal diastolic function. In vitro studies examining myocyte relaxation time and contractility have identified IL-1β, IFN-γ, and poly(I:C) as inducers of abnormal cardiomyocyte relaxation time without affecting contractility (175). In vivo, inhibition of IL-1β or macrophage depletion ameliorated diastolic dysfunction in a HFD model of HFpEF, whereas infusion of IL-6 alone can induce cardiac hypertrophy, fibrosis, and diastolic dysfunction (155, 176). Although targeting individual cytokines in those who have already developed overt HFpEF has proven unsuccessful in clinical trials, a preventive strategy as seen in CANTOS may prove fruitful. The challenge, however, remains in identifying high-risk populations that may benefit from this strategy.

Cardiomyocyte hypertrophy is another important contributor to the HFpEF phenotype, and inflammation, via immune cell-cardiomyocyte interaction and secreted factors, is a key driver of LV hypertrophic remodeling. In mice exposed to increased afterload, macrophage-specific knockout of microRNA-155 reduced monocyte/macrophage myocardial infiltration, cardiomyocyte hypertrophy, and cardiac function (177). Innate and adaptive immune cell cross talk can also modulate LV hypertrophic remodeling. Suppression of CCR2⁺ macrophages, for example, attenuates lymph node CD3⁺ T cell expansion and cardiac hypertrophy (178). Interestingly, CXCL1-CXCR2 signaling has been implicated in myocardial monocyte infiltration and hypertrophic remodeling in ANG II-infused animals and may mediate the myocardial accumulation of inflammatory cells in HFpEF (179). Although the specific cell sources have not been identified, individual cytokines, such as IL-6, IL-1β, TNF-α, and TGF-β, have also been implicated in cardiomyocyte hypertrophy (180–182).

Adipoinflammatory Myocardial Remodeling

As discussed previously, AT remodeling is an important contributor to systemic and myocardial inflammation in cardiometabolic HFpEF. Adipocyte-immune cell cross talk is well characterized in obesity-induced AT inflammation and is beginning to emerge as an important player in myocardial remodeling.

Adipocyte-secreted free fatty acids trigger resident macrophage release of TNF-α, establishing a paracrine loop that worsens AT inflammation (183). Such cross talk can also act to promote myocardial remodeling as seen with a recently identified macrophage-adipocyte axis; macrophages recruited to the myocardium in HF secrete urokinase plasminogen activator (uPA), cleaving adipocyte-secreted full-length platelet-derived growth factor-D (PDGF-D ) and promoting pathologic myocardial remodeling (184, 185).

Adaptive immune cell-AT interactions have also been reported in CVD. FALCs, harboring B and T cells, are present in EAT and can expand in obesity and inflammation to mediate systemic and myocardial inflammation in pathological states (48). In mice, B-cell depletion from FALCs results in reduced bone marrow granulopoiesis and neutrophil infiltration to the myocardium after myocardial infarction, reducing fibrosis (42). Similarly, global B cell depletion in ANG II-infused mice using a CD22 antibody resulted in reduced myocardial remodeling and improved function (186). These clusters can also interact with myeloid cells potentially modulating their activity (48, 187). Such interactions are yet to be reported in HFpEF; furthermore, their existence is sporadic. These findings highlight the role of EAT as not only a source of inflammatory adipokines but an immune cell reservoir directly affecting the myocardium.

Exacerbation of EAT inflammation in obesity via ACE2 knockout results in worsened cardiac function and steatosis (47). Similarly, activation of hypoxia-inducible factor 2α (HIF2α ) in AT results in pathological cardiac hypertrophy, elevated levels of circulating proinflammatory cytokines, and myocardial inflammation via NF-κB and nuclear factor of activated T-cell (NFAT) activation (188). Myocardial homeostasis is directly affected via an array of secreted factors including inflammatory adipokines. Secreted retinol-binding protein 4 (RBP4) induces the expression of TLR4 and myeloid differentiation primary response gene 88 (MyD88) resulting in cardiomyocyte inflammation, hypertrophy, and oxidative stress. RBP4 also reduces the expression of glucose-transporter 4 (GLUT4) and mediates cardiomyocyte insulin resistance, a cardinal feature of cardiometabolic HFpEF (189).

AT also mediates protective homeostatic functions under normal conditions. Cross talk between brown AT (BAT) and the myocardium can prevent hypertension-induced cardiac hypertrophy via adenosine A2A receptor-derived secretion of FGF21 from BAT (190). Similarly, adiponectin is a cardioprotective adipokine, whose levels decrease in obesity (191). Reversal of AT inflammation and promoting BAT formation may prove productive in reducing pathologic changes seen in cardiometabolic HFpEF.

Inflammation-Mediated Metabolic and Energetic Dysfunction

Structural and bioenergetic abnormalities in cardiomyocyte mitochondria are characteristic of HFpEF, resulting in a mismatch between ATP demand/production and energy deficit (192). Mitochondrial dysfunction is accompanied by altered cardiomyocyte metabolism, including metabolic inflexibility, impaired fatty acid and glucose oxidation, and increased fatty acid uptake with myocardial steatosis (85, 152). ROS have been proposed as the major contributor to mitochondrial dysfunction and can promote inflammation, cardiac remodeling, and diastolic dysfunction (193, 194). However, evidence is emerging that implicates inflammation-induced mitochondrial dysfunction as an important driver of HFpEF.

Mitochondrial protein acetylation is a crucial posttranslational modification regulating mitochondrial energy metabolism. Mitochondrial hyperacetylation in HFpEF occurs via sirtuin 3 downregulation and a defective NAD⁺ salvage pathway and is linked to impairment of oxidative metabolism and metabolic inflexibility, both hallmarks of HFpEF (85, 195). Evidence is emerging that implicates hyperacetylation as a driver of inflammation. Mitochondrial hyperacetylation increases NLRP3 assembly triggering increased production of IL-1β/IL-18 and resulting in mitochondrial dysfunction in cardiometabolic HFpEF. Interestingly, β-hydroxybutyrate supplementation reverses diastolic dysfunction, NLRP3 activation, and mitochondrial dysfunction by reducing mitochondrial hyperacetylation (196).

The proinflammatory cytokines IL-6, IL-1β, TNF-α, and TGF-β have all been implicated in mitochondrial dysfunction via ROS generation (154–156). Similarly, iNOS-mediated nitrosative stress triggers mitochondrial dysfunction in HFpEF. iNOS inhibition ameliorated mitochondrial injury and dysfunction while also reducing Akt S-nitrosylation and thereby insulin resistance in HFpEF (151). Mitochondria can also contribute to inflammation via ROS generation and the release of DAMPs that trigger immune cell activation (197). This generates a cycle of inflammation and mitochondrial dysfunction. Thus, targeting inflammatory pathways may prove successful in ameliorating mitochondrial dysfunction in HFpEF.

Conclusions and Perspectives

In light of important experimental and clinical findings discussed here, targeting inflammation as a therapeutic strategy in HFpEF warrants further investigation. Whereas the DHART2 trial revealed reductions in NT-proBNP and high-sensitivity CRP with IL-1β blockade in HFpEF, it did not translate into improvement in clinical parameters (198). A large-scale, randomized clinical trial may be required to derive reliable conclusions regarding IL-1β blockade in HFpEF; in the meantime, reliance on alternate strategies that target inflammation differently or more broadly may prove fruitful. Examining anti-inflammatory mechanisms as a preventive strategy in a high-risk subset of patients may also prove successful and warrants further consideration. Also, worthy of further exploration is delineation of subsets of HFpEF patients manifesting strong inflammatory profiles who might benefit from anti-inflammatory therapies.

In summary, comorbidity-driven inflammation is emerging as a critical contributor to HFpEF pathogenesis and progression. Cardiomyocyte dysfunction and pathologic ECM remodeling occur as downstream effects of chronic low-grade inflammation, generating myocardial stiffness and reduced ventricular compliance. Discovery of novel inflammatory mechanisms driving key pathophysiological events in HFpEF is paramount to better understanding of syndrome development and progression and therapeutic intervention.