Inflammation is a physiological defence mechanism of the body against injurious stimuli such as tissue damage and infection. Timely inflammation in adequate intensity is essential to eliminate harmful stimuli; an insufficient inflammatory response can result in persistence of the trigger. Active resolution of inflammation is also essential as it facilitates tissue healing after injury; failure to resolve leads to chronic inflammation, extended tissue destruction and progressive fibrosis [5, 6]. Inflammation and fibrosis can thus be viewed as a continuum of events within the framework of tissue defence, repair and regeneration.

The inflammatory response is extremely complex and comprises several stages including vascular phase, cellular phase and resolution phase. Leukocytes are major cellular effectors that direct this response through various mechanisms, including chemical mediators such as cytokines [7]. During the inflammatory process, the endothelial layer undergoing activation and selective changes in permeability allows cellular components to shift from intravascular to extravascular compartment [8]. Secreted proteins and extracellular matrix (ECM) components also play a vital role in inflammation by directly moderating the inflammatory cascade or by providing signals to cellular components of inflammation. Osteopontin, a phosphorylated glycoprotein secreted by monocytes and lymphocytes, mediates leukocyte adherence and migration [9]. Versican is an ECM proteoglycan, also involved in leukocyte adherence and migration; it is abundantly expressed and produced by activated macrophages and stromal cells during inflammation [10]. Hyaluronic acid (HA) is a glycosaminoglycan ECM component having a dual role in inflammation. While native polymeric HA is typically anti-inflammatory [11, 12], the smaller fragments elicit a pro-inflammatory response by binding to toll-like receptor 2 (TLR2) and TLR4 of monocytes, dendritic cells and lymphocytes [13]; TLRs are a class of proteins that play a key role in the innate immune system. Recent studies also indicate that low molecular weight HA fragments promote a classically activated “pro-inflammatory” state in macrophages [14].

Resolution of inflammation is an active process orchestrated by “pro-resolution” factors. These factors induce “pro-resolution” programmes in stromal cells and provide cues to inflammatory cells such as neutrophils to undergo apoptosis. They also enhance efferocytosis and later signal macrophages to exit via lymphatic vessels [6, 15••]. Polyunsaturated fatty acid-derived resolvins and protectins function as proresolution factors and play a key role in subduing inflammation [16, 17] (Fig. 1). Inflammation is further modulated by a number of checkpoints. For instance, TLR-mediated inflammasome activation is countered by a negative internal feedback mechanism involving phosphoinositide 3-kinase (PI3K) and excessive TLR signalling is moderated by negative regulators of immune responses, such as interleukin-1 receptor-associated kinase-M (IRAK-M) and suppressor of cytokine signalling-1 (SOCS-1). T-regulatory cells also actively inhibit inflammation by producing several anti-inflammatory cytokines [18, 19]. Failure of such regulatory mechanisms could lead to a state of chronic inflammation causing continuous tissue damage and progressive fibrosis.

Fibrosis is an essential component of tissue repair that follows tissue injury and is usually associated with inflammation. The aim of fibrosis is to deposit connective tissue in order to preserve tissue architecture; progressive fibrosis reflects a pathologic state and results in scarring, impairment of function and organ damage [5, 20].

Myofibroblasts are major cells responsible for ECM secretion; they arise directly from fibroblasts or from other cell types such as macrophages, endothelial cells, pericytes and circulating monocytes. Several literature reviews exclusively discuss the role of (myo)fibroblasts in fibrosis and interested readers are directed to them [21, 22•].

Macrophages also play a pivotal role in secretion of ECM components and in ECM remodelling. They are major sources of matrix metalloproteinases (MMPs) and tissue inhibitor of metalloproteinases (TIMPs) [23] and are the primary cells involved in the phagocytosis of cellular debris and infectious agents. The phagocytosed particle can influence phenotypic characteristics of macrophages [24, 25]; for instance, macrophages assume a more fibrotic (M2) phenotype after ingesting apoptotic neutrophils [26]. Cytokines such as interleukin-13 (IL13) and IL4 also induce profibrotic (M2) phenotypic changes in naïve (M0) macrophages. M2 phenotype is characterized by reduced expression and secretion of inflammatory mediators, e.g. tumour necrosis factor-α (TNFα) and IL6, and augmentation of cell survival and fibrotic signals, e.g. IL10, insulin-like growth factor-1 (IGF1), transforming growth factor-β (TGFβ) and galectin-3 (Gal-3) [27•, 28•]. Besides promoting fibrosis, M2 macrophages also endocytose collagen utilizing mannose receptors highlighting their pleiotropic role in ECM homeostasis [29]. Other immune cells, e.g. neutrophils, lymphocytes and eosinophils, also contribute to the development of fibrosis in various organs [7]. Extensive communication between inflammatory cells, fibroblasts and ECM actively modulates the fibrotic response [30‐33].

a. Immune cells as a source of cardiac inflammation: Neutrophils and monocytes home to the site of cardiac injury and release aggressive mediators such as reactive oxygen species (ROS) and proteases, with the primary aim of eliminating the factors that caused the cardiac insult. However, this nonspecific response could also result in extensive damage to the healthy cardiac tissue [35]. Macrophages exposed to inflammatory signals, e.g. Interferon-γ (IFNγ) and IL1, typically assume a pro-inflammatory M1 phenotype [36]. These macrophages sustain cardiac inflammation by secreting inflammatory cytokines themselves and can also signal neighbouring fibroblasts and cardiomyocytes to adopt pro-inflammatory phenotypes [37]. In subsequent stages of inflammation, effectors of innate immune system are modulated by lymphocytes; for instance, cytokines secreted from Th1 cells sustain inflammation while Th2 cytokines produce anti-inflammatory and prohealing signals [38].

b. Pro-inflammatory cardiomyocytes in cardiac injury: Cardiomyocytes (~30–40% of cells in the healthy heart) secrete pro-inflammatory cytokines typically after hypoxia or cardiac injury [39•]. TNFα expression is upregulated in hypoxic cardiomyocytes [40] while lipopolysaccharide (LPS) stimulation of cardiomyocytes in vitro increases IL6 production [39•]. IL6 and other related cytokines secreted by cardiomyocytes are pivotal in regulating cardiac myocyte hypertrophy and apoptosis [41]. Moreover, IL6 is known to direct the nature of inflammation from acute to chronic, by changing the leukocyte infiltrate from neutrophils to monocyte/macrophages [42].

c. Cardiac fibroblasts as a source of pro-inflammatory cytokines: Although mentioned frequently in the context of fibrosis, cardiac fibroblasts exposed to an inflammatory milieu, e.g. TNFα, transform to a pro-inflammatory phenotype, with increased expression of cytokines such as IL1β and IL6 [43]. When activated by mechanical stress, they produce pro-inflammatory mediators such as monocyte chemoattractant protein-1 (MCP-1), IL8 and biglycan [44••]. Cardiac fibroblasts also sustain and perpetuate pre-existing inflammation; they directly facilitate transendothelial migration of leukocytes by producing gelatinases such as MMP9. Co-culture of fibroblasts with macrophages also increases macrophage inflammatory protein-1α (MIP-1α) expression in macrophages and enhances reciprocal enhancement of monocyte-fibroblast adhesion and chemokine production [30]. Furthermore, cardiac fibroblasts stimulated by IL-17A produce chemokines such as MCP-1, IL6 and leukaemia inhibitory factor (LIF), responsible for recruiting and differentiating myeloid cells, and this mechanism has been implicated in the pathophysiology of inflammatory dilated cardiomyopathy (IDCM) [45].

TLRs are a part of the innate immune system and play a crucial role in the development of inflammatory disorders by initiating both innate and adaptive immune responses. They are essentially pattern recognition receptors (PRRs) designated to recognize infectious or dangerous foreign patterns collectively termed as pathogen-associated molecular patterns (PAMPs) and danger-associated molecular patterns (DAMPs) [46]. TLR4 is usually expressed in monocyte-macrophage-lineage cells also in fibroblasts and epithelial cells. Recent work by Liu and colleagues demonstrate upregulation of TLR4 in cardiomyocytes in HF [47••]. Lipid A component is an important exogenous ligand for TLR4, while various intracellular and extracellular components (e.g. heat shock proteins (HSP), fibrinogen, heparin sulphate, HA) serve as endogenous ligands [48]. Intracellular TLR4 signalling can occur via both the myeloid differentiation primary response gene-88 (MyD88)-dependent pathway resulting in early nuclear factor-κβ (NFκβ) activation or the MyD88-independent pathway resulting in late NFκβ activation [49].

TNF-NFκβ pathways are indicated in cardiac infection and injury, while viral triggers typically activate retinoic acid-inducible gene-1 (RIG-1) pathways. Other cardiac inflammatory mechanisms include caspase-1-inflammasome pathways, activated usually during oxidative and cellular stress [50]. Persistent activation of various cardiac inflammatory pathways could serve as a precursor to fibrotic changes, resulting in pathological remodelling of the heart.

MI usually occurs after a vascular insult to the myocardium and is characterized by extensive necrosis of cardiomyocytes. This results in leakage of intracellular contents and accumulation of ROS. The released DAMPs together with cytokine signals from the neighbouring cells constitute the “alarmin-response” [83]. However, the infarcted area has limited or no vascularization, and this prevents the blood-borne immune cells from gaining immediate access to the necrotic core. During these initial stages of ischaemic damage, cardiac myofibroblasts could potentially take over the phagocytic role of macrophages, by actively engulfing dead cells [84].

This is followed by an intense and transient inflammatory phase, characterized by a “neutrophil-monocyte” infiltration [85]. However, both resident cardiac cells (cardiomyocytes, cardiac fibroblasts, resident macrophages, mast cells) and recruited cells (leukocytes) contribute to the development of sterile inflammation post-MI [85, 86•]. The innate immune cells recognize the released alarmins utilizing TLRs and activate downstream inflammatory pathways, and TLR2 and TLR4 are crucial players in the post-infarct inflammatory reaction.

Inflammation is further sustained by upregulation of various pro-inflammatory cytokines, e.g. MCP-1, TNFα and IL6, within the infarcted myocardium. MCP-1 is involved in the recruitment of monocytes while TNFα enhances adhesion and extravasation of leukocytes through the endothelium [87‐89]. TNFα is an acute-phase protein involved in both post-MI inflammatory reaction and ischaemia-reperfusion (I/R) injury [90, 91•]. The role of IL6 in cardiac inflammation and remodelling is ambiguous. Enhanced IL6 expression could accentuate the inflammatory response and exacerbate the deleterious after-effects of MI [42, 56]. However, knocking out IL6 confers no protective effect in a mouse model of MI [92]. Moreover, IL-6 receptor inhibition did not improve cardiac function after I/R in a recent study [93]. There are also changes in ECM around the necrotic area after MI. For instance, large polymers of HA are degraded to low molecular weight HA and together with fibronectin fragments initiate and sustain a multitude of inflammatory cascades [94].

Molecular stop signals of inflammation such as IRAK-M in macrophages and fibroblasts actively wean the post-MI inflammatory response. They prevent uncontrolled TLR/IL1-mediated responses by acting as a functional decoy to attenuate sustained inflammatory response and improve adverse post-infarction cardiac remodelling [95].

In the proliferative phase that follows, macrophages secrete several cytokine growth factors and activate mesenchymal reparative cells to deposit ECM [85]; Gal-3, a profibrotic protein produced predominantly by macrophages, is a major player in post-MI cardiac remodelling [27•, 77••, 96]. TGFβ, another key fibrotic cytokine, aids in repair by supressing inflammation and stimulating hypertrophic cardiomyocyte growth after MI. TGFβ also promotes ECM deposition by upregulating collagen and fibronectin synthesis and downregulating ECM degradation [28•, 97]. Crosstalk between M2 macrophages and fibroblasts together with Th2 responses sustains the fibrotic response. Recent studies also suggest the indispensable role of proteoglycans such as syndecan-1 and 4 in post-MI remodelling and fibrosis of the heart. Although mice lacking syndecan-1 and 4 showed marked reduction in profibrotic signalling, this resulted in increased cardiac rupture after MI [80, 81].

Apoptosis of the majority of reparative cells marks the end of the proliferative stage and infarct maturation occurs with the formation of cross-linked collagen. The extent of post-MI remodelling depends on the infarct size and the quality of cardiac repair. The infarct zone undergoes replacement fibrosis while the surrounding non-infarct zone displays perivascular and interstitial fibrosis [98•]. The aim of the fibrotic response is to preserve structural integrity and to maintain the pump function of the heart by preventing dilatation, aneurysm formation or myocardial rupture [99]. However, failure of cardiac myofibroblasts to undergo apoptosis or persistence of profibrotic signalling could result in pathological remodelling of the heart.

Viral infection is a common cause of myocarditis and is characterized by inflammation of the myocardium; we discuss the sequence of events from infection to fibrosis in group B Coxsackie viral (CVB) infection.

Macrophages and lymphocytes of Peyer’s patches and the spleen serve as ports of entry for CVB3 viral particles, and they reach the heart through the bloodstream. Utilizing endothelial receptor CAR (coxsackievirus and adenovirus receptor), primarily located in the intercalated discs of the adult heart or receptor DAF (delay accelerating factor), they translocate into cardiomyocytes [100]. CAR-deficient mice are resistant to both cardiac infection and inflammation, clearly suggesting that in the acute phase of myocarditis, most of the damage is mediated by the virus. Lindner et al. demonstrated that when cardiomyocytes and cardiac fibroblasts were both infected with CVB3, cardiac fibroblasts displayed a tenfold increase in viral replication, indicating their crucial role in contributing to the viral load in myocarditis [101••].

TLR3 is involved in viral recognition and in mounting antiviral type II interferon response; mice lacking TLR3 developed severe viral myocarditis highlighting the protective action of this TLR in CVB3 infection [102]. After entering the cardiomyocytes, the viral machinery is actively replicated. Viral proteases such as enteroviral protease-2A cleave dystrophin and dystrophin-associated glycoproteins [103]. This could result in the loss of tethering of the cardiomyocytes to the ECM, leading to cardiomyocyte-ECM uncoupling [104]. Subsequent cardiomyocyte loss occurs via necrosis or apoptosis and is usually followed by replacement fibrosis. The viral PAMPs and released cellular contents are also recognized by other TLRs, and this leads to activation of other pro-inflammatory cascades [100]. The role of inflammation-induced damage in the acute phase is demonstrated by the fact that TLR4-deficient mice were protected against CVB-induced cardiac injury [105, 106••].

Although hypertension has a strong genetic component, neurohormonal activation, oxidative stress and low-grade systemic inflammation play a vital role in its aetiology, especially in insulin-resistant states. Hypertension is a leading cause of HF and exerts a deleterious effect on the cardiovascular system through direct haemodynamic mechanisms and also through overactivation of the renin-angiotensin-aldosterone system (RAAS) [110].

Hemodynamic parameters such as increased shear stress together with low-grade systemic inflammation promotes endothelial damage in hypertension. During the course of time, this manifests itself as perivascular fibrosis with considerable deposition of collagen in the adventitia of intramural arteries, resulting in reduced vascular compliance and changes in permeability. Hypertension also elicits structural and functional changes in microcirculation leading to microvascular remodelling and rarefaction [111].

There are also simultaneous changes in the cardiac tissue; progressive deposition of collagen in cardiac ECM results in reactive interstitial fibrosis. Although this develops without cardiomyocyte loss, it decreases myocardial compliance and clinically manifests as HF with preserved ejection fraction (HFpEF) [112]. In advanced hypertension, there is a pathological hypertrophy of cardiomyocytes and also an increased loss of cardiomyocytes. This results in irreversible replacement fibrosis leading to deterioration of the systolic function of the heart, clinically manifesting as HF with reduced ejection fraction (HFrEF) [51, 110, 113]. In animal models of sudden pressure overload (e.g. TAC), the results are more dramatic with accelerated cardiomyocyte loss and more rapid onset of cardiac fibrosis [114].

RAAS is the key homeostatic hormonal mechanism that maintains blood pressure in order to ensure adequate tissue perfusion. However, stimulation of the RAAS also elicits pro-inflammatory and profibrotic responses and contributes to cardiovascular remodelling. For instance, aldosterone has been implicated in the development of cardiac fibrosis in hypertension [115]; renin overexpression in hypertensive rats leads to cardiac remodelling and diastolic dysfunction via a fibrosis-independent “titin-related” mechanism [116].

As Ang II is the key vasoconstrictive protein in this axis, we briefly discuss few novel mechanisms of Ang II-related cardiovascular remodelling. Ang II can act both independently and via the classic TGFβ axis to induce fibrosis [117]. Recent studies describe the Ang II-Gal-3-IL6 axis as a modifiable fibrotic pathway in hypertension. Genetic inhibition of IL6 resulted in reduction of cardiac inflammation and fibrosis in an Ang II high-salt-induced hypertension mouse model. IL6 deletion also improved cardiac dysfunction although there was no net reduction in blood pressure [118••], suggesting the critical role of IL6 in the mediation of cardiac inflammatory and fibrotic effects of Ang II. Subsequent studies in models of chronic Ang II-induced hypertension demonstrated that genetic ablation of Gal-3 also reduced myocardial macrophage infiltration and fibrosis, highlighting the causative role of Gal-3 in cardiac fibrosis related to hypertension [119]. ECM proteins such as osteopontin are also involved in Ang II-induced cardiac fibrosis, and studies with osteopontin^−/− mice indicate that there is a significant reduction in cardiac fibrosis after 3 weeks of Ang II infusion [120]. Other experimental studies, also with murine models, suggest that syndecan-1 amplifies profibrotic effects of Ang II and is a critical regulator of fibrosis in the heart [82]. Recently, neutrophil-generated S100a8/S100a9 proteins have been implicated in Ang II-induced cardiac inflammation and fibrosis [121]. There is also accumulating evidence on the role of cardiac mast cell-IL4 axis in the mediation and development of hypertension-related cardiac fibrosis [63]. Targeting these novel pathways of inflammation and fibrosis could effectively prevent or reduce cardiovascular fibrosis in the setting of hypertension.

Although most patients survive the primary cardiac event due to early detection and timely management, every cardiac insult decreases the cardiac contractile reserve and these patients have an increased risk of developing HF [122]. HF can be defined as the inability of the heart to adequately maintain cellular perfusion under normal cardiac filling pressure. While half of the patients with HF exhibit decreased ejection fraction (HFrEF), the other half have a normal EF (HFpEF). Based on clinical presentation, HF can be classified as acute HF (AHF), when the patient presents with cardiac decompensation, and CHF, when the patient has impaired cardiac function but is compensated and stable, i.e. able to maintain tissue perfusion without assistance [123, 124].

AHF is characterized by a systemic inflammatory response, with elevated pro-inflammatory cytokines [125]. Other triggers may be present that provoke inflammation: AHF is often accompanied by viral or bacterial infection and is usually preceded by MI or atrial fibrillation (AF). However, after the acute event has been treated, a chronic response develops, and such patients frequently develop CHF. Other concomitant factors such as hypertension can also contribute to the development of AHF and CHF. It is also important to note that CHF can itself be a predisposing factor for the development of future AHF, and the goal of CHF management is to maintain the patient in compensated HF state and prevent them from deteriorating into a state of acute decompensated HF (ADHF) [124].

Inflammation in the setting of CHF can be very complex. Low-grade systemic inflammation can both be a cause and consequence of HF [34, 126‐128]. Moreover, chronic oxidative stress associated with HF can exacerbate the pre-existing inflammatory state [129]. It is hypothesised that the presence of co-morbidities might lead to increased inflammation and to HF [130•]. There is also upregulation of TLR4 in cardiomyocytes during HF, suggesting the direct role of cardiomyocytes in cardiac inflammation associated with HF [47••]. Sustained activation of protective neurohormonal mechanisms can also contribute to ongoing cardiac inflammation and fibrosis [131, 132••], resulting in further loss of cardiac function and clinical deterioration up to the point of ADHF and death.

Thus, it is highly relevant to address this question in patients with CHF: Is cardiac fibrosis progressive? Serial biomarker measurements and imaging modalities such as cardiac magnetic resonance imaging (CMR) can help us answer this question by aiding in identifying patients with progressive cardiac fibrosis [133‐135] (Fig. 3).