Introduction

Lymphatic vessels play a key role in maintenance of tissue homeostasis by returning extracellular fluids, proteins and solvents, but also immune cells and lipids, back to the blood circulation. Recent research has revealed that lymphatic vessel remodeling, through the process of lymphangiogenesis and/or altered lymphatic function, occurs in many different diseases (1). When lymphatic remodeling is insufficient or maladaptive, it causes lymphatic transport dysfunction, contributing to installation of chronic interstitial edema and inflammation that triggers development of interstitial fibrosis. Noteworthy examples include patients with secondary (acquired) lymphedema, such as after mastectomy, where the affected limb progresses from reduced lymphatic transport to clinical edema with local inflammation and eventually fibroadipose tissue deposition (2). In the context of cardiac diseases, myocardial interstitial fibrosis is one of the hallmarks of myocardial remodeling, contributing to both diastolic and systolic dysfunction and to the development of heart failure (HF), having a negative impact on patient prognosis (3). However, most therapies currently used in HF patients have failed to exert a robust antifibrotic effect. Similarly, there are no approved medical treatments to improve lymphatic transport. Promisingly, therapeutic lymphangiogenesis has been shown experimentally to limit edema and/or inflammation in different pathologies such as secondary lymphedema, skin burn, irritable bowel syndrome, and nephropathy (1,4,5), as well as in cardiovascular diseases including atherosclerosis and myocardial infarction (MI) (6,7).

In this review, we focus on presenting the links between lymphatics, edema, inflammation, and interstitial fibrosis in the context of cardiac pathology.

Cardiac Lymphatics and Edema in Cardiac Disease

The heart has an extensive lymphatic network that drains extracellular fluid and waste from the myocardium with each cardiac contraction, propagating lymph in collecting lymphatics, from the apex to the base, toward draining lymph nodes situated around the aortic arch (8). The lymphatic flow rate is affected by the force of cardiac contractions and the duration of diastole. Central venous pressure is another major determinant of lymphatic drainage, because it limits lymph return centrally (8). Further, changes in tissue interstitial fluid pressure (IFP) impact lymphatic function. Indeed, lymphatic endothelial cells (LECs) of capillary lymphatics are attached, not to a basement membrane like blood vessels, but to surrounding extracellular matrix (ECM) components, such as collagen I and fibronectin, through elastic anchoring filaments composed of fibrillin and emilin-1. During edema, the increased pull by distorted ECM components on anchoring filaments allows lymphatic capillaries to remain open and permeable to fluid and solvents to ensure increased drainage. Indeed, studies in dogs have revealed that the heart responds to acute myocardial edema induced by myocardial ischemia, arterial hypertension, or cardioplegia with a robust increase in cardiac lymphatic transport as a compensatory mechanism (9). However, during inflammation, infiltrating neutrophils release enzymes, such as neutrophil elastase, that degrade emilin-1 anchoring filaments, leading to weakening of intercellular junctions and lymphatic vessel collapse (10), and as a result poor lymphatic transport and installation of edema. Given the observation that many cardiovascular diseases are accompanied by chronic interstitial edema, lymphatic transport capacity may be insufficient to balance vascular hyperpermeability-induced extravasation in pathology.

Alteration of lymphatic transport may be induced by insufficient or maladaptive lymphatic remodeling. In adults, lymphatic vessels are quiescent, but lymphangiogenesis is reactivated in many pathological conditions. Recent studies in humans and rodents have revealed that cardiac lymphatics remodel in cardiovascular diseases (reviewed in Brakenhielm and Alitalo [11]) (Figure 1). Severe alterations, including dilation of lymphatic capillaries, have been found in ischemic heart disease (post-MI) and in terminal stages of HF (11–14). However, this lymphatic remodeling included expansion of lymphatic capillaries but slimming of lymphatic pre-collectors, associated with lymphatic transport dysfunction (6). Interestingly, murine studies have shown that cardiac lymphangiogenesis occurs in both infarct zone and bordering noninfarcted myocardium (6,14). Lymphatic remodeling has also been reported in patients with cardiac inflammation, such as endocarditis and myocarditis (13,15), and after cardiac transplantation (12), as well as in other chronic conditions such as severe aortic stenosis (13) or hypertrophic cardiomyopathy (16).

Insufficient lymphangiogenesis and/or lymphatic dysfunction has been shown to aggravate myocardial edema and inflammation post-MI in rodents (6,17,18). Of note, clinical studies, based on cardiac magnetic resonance imaging (CMR), have revealed that myocardial edema is a common denominator in many cardiovascular diseases (9,19), including MI (20), myocarditis (21), acute decompensated HF (22), nonischemic dilated cardiomyopathy (23), cardiac amyloidosis (24), and takotsubo cardiomyopathy (25). Similarly, in patients with pulmonary arterial hypertension pericardial effusion has been reported to be associated with elevated mortality risk (26). Of note, noncardiac systemic factors such as serum hypoalbuminemia, commonly found in elderly HF patients due to liver dysfunction or nephrotic syndrome, may also contribute to myocardial edema in HF by reducing plasma oncotic pressure and inducing a fluid shift from the intravascular to the interstitial space (27).

Importantly, interstitial myocardial edema increases IFP both in acute and chronic conditions, leading to decreased contractility (28) and increased left ventricular stiffness (29), having an overall negative impact on cardiac function (9,30,31). Of note, a relatively small change in interstitial pressure can have relevant effects on cardiac output (30). Additionally, chronic lymphatic blockade leads to myocardial interstitial fibrosis, which also increases left ventricular stiffness and interstitial resistance contributing to diastolic dysfunction (30,31). Moreover, in amyloidosis patients, myocardial edema was independently associated with a higher mortality (24). Thus, because reducing overall congestion in HF patients improves quality of life and reduced cardiovascular events (32), it seems likely that reducing myocardial edema might also be beneficial. Of note, in patients with advanced decompensated HF, hemodynamic unloading with vasodilators and diuretics was associated with a relief of CMR-assessed myocardial edema (22). Their impact in chronic conditions deserves further studies. In parallel, improvement of lymphatic uptake by stimulation of lymphangiogenesis represents an interesting novel approach.

Regulation of Lymphangiogenesis

Vascular endothelial growth factor (VEGF) family members VEGF-C and VEGF-D are the most potent and specific prolymphangiogenic factors, signaling through vascular endothelial growth factor receptor 3 (VEGFR3) (1). VEGF-C is constitutively expressed at low levels in most tissues by several cell types, including cardiomyocytes. VEGF-C levels increase during inflammation (33). In addition, other stimuli, such as hyperosmolarity (34), and/or increased IFP or flow rates (35,36), may also augment VEGF-C production. Besides stimulating lymphangiogenesis, VEGF-C may also modulate local ECM by altering expression of extracellular enzymes, such as matrix metalloproteinases (MMPs) and heparanases, in lymphatics. Indeed, VEGF-C promotes formation of heparane sulphate (HS)-anchored chemokine gradients regulating lymphocyte migration from tissues to lymphatics (37).

Tissue edema may further induce lymphangiogenesis through the release of mediators from the ECM, such as the glycosaminoglycan hyaluronan, which is degraded by hyaluronidases into bioactive prolymphangiogenic fragments acting on the lymphatic vessel endothelial hyaluronan receptor (LYVE)-1 to stimulate LEC proliferation and migration in synergy with VEGF-C (38). In addition, interstitial fluid accumulation, which increases IFP, notably in the heart (9), has an impact on the organization of ECM components, including fibronectin that signals to LECs via β1 integrins directly associating with VEGFR3 to amplify VEGF-C signaling (39). Conversely, whereas quiescent LECs weakly express α4β1, VEGF-C stimulation up-regulates this integrin (40). In vitro studies have further confirmed that cell stretch synergizes with VEGF-C in stimulation of LEC proliferation (41). Interestingly, in human LECs, mechanical stretch was recently shown to lead to dissociation of integrin-linked kinase (ILK) from β1 integrin, thus freeing this integrin for interactions with VEGFR3 (39). Similarly, in vitro exposure to VEGF-A induced the collagen receptors integrin α1β1 and α2β1 in LECs, whereas antibodies against α1β1 and α2β1 reduced wound healing lymphangiogenesis in mice (42), highlighting the important role of ECM-integrin signaling in regulation of lymphatic remodeling. However, given that edema often is accompanied by inflammation, known to adversely affect LEC junctions and barrier function (43,44), the links between increased IFP and functional lymphangiogenesis is complex and depends on the local microenvironment.

Linking Cardiac Lymphatics, Edema, and Fibrosis

Early studies of selective chronic myocardial edema, induced by lymphatic ligation in dogs, demonstrated endomyocardial fibrosis and endocardial fibroelastosis (45). This was confirmed in more recent studies in rabbits where cardiac lymphatic ligation-induced edema rapidly increased cardiac collagen types I and III production (46). Chronic myocardial edema, induced by surgical pulmonary arterial hypertension in rats, led to transiently increased production of collagen type I in the left ventricle, and to a persistent increase in the expression of collagen type III and prolyl 4-hydroxylase (P4H), an enzyme involved in the formation of collagen fibrils (47). By contrast, collagenase activity was decreased, and as a consequence, collagen content progressively increased over the 1-month study. Similarly, in dogs, chronic myocardial edema, induced by arterial hypertension and right heart pressure elevation, increased myocardial collagen content in the absence of any hypertrophic or inflammatory stimuli (30,31).

Although the available clinical data are scarce, lymphatic capillary density was found to be correlated with the extent of myocardial interstitial fibrosis in patients with hypertrophic cardiomyopathy (16). Additionally, autopsy studies in humans showed that in different cardiac pathological conditions lymphatic remodeling and expansion accompanies chronic inflammation and interstitial fibrosis (13).

Intriguingly, studies in dogs have suggested that the switch in collagen isoforms induced by chronic edema, from stiffer collagen type I to more compliant collagen type III, may serve to safeguard against loss of cardiac compliance during episodes of acute edema (30,31). In response to increased coronary sinus pressures, in healthy dogs, there was acute edema accompanied by a large increase in IFP and a severe drop in cardiac compliance (Figure 2). However, in dogs with chronic myocardial edema, induced by pulmonary banding, the acute aggravation of edema, induced by coronary sinus occlusion, was followed by a similar increase in IFP as in healthy dogs, but compliance was significantly less reduced, indicating resilience of the remodeled ECM to acute pressure changes.

Edema can activate profibrogenic mechanisms in fibroblasts. Studies in cardiac explants from pigs have indicated physiological IFP in the range of 5 to 60 mm Hg with increases up to 200 mm Hg during edema induced by ischemia–reperfusion (48). Studies in dogs similarly confirm a steep curve between cardiac water content and IFP in the ventricular wall (Figure 2), although the absolute values reported were ∼15 mm Hg in physiology and ∼50 mm Hg during acute edema (9,31). Such important pressure changes can be sensed by cells firmly attached to the ECM, including LECs and fibroblasts. Indeed, in response to increased mechanical stress within the ECM, fibroblasts differentiate towards the profibrogenic myofibroblast phenotype, produce more collagen, and increase the stiffness of the ECM (49).

Additionally, fibroblasts may respond to changes in interstitial flow (50). Physiological interstitial flow is the movement of fluid through the ECM, often between blood vessels and lymphatic capillaries, providing a specific mechanical “shear stress” to interstitial cells. In vitro studies have revealed that increased interstitial flow increased collagen type III, α-smooth muscle actin (a marker of myofibroblast differentiation), and TGF-β₁ expression in cardiac fibroblasts seeded in 3-dimensional collagen gels, in a mechanism involving angiotensin II/AT₁ receptor signaling (51). Therefore, changes in interstitial flow velocity and hydrostatic pressure associated with edema may influence fibroblast activity.

Interstitial pressure can be transmitted to fibroblasts through transmembrane receptors such as syndecan-4, a heparan sulfate proteoglycan acting as a transmitter of fibroblast mechanical stress, to increase collagen deposition and strengthen the ECM. Syndecan-4 is up-regulated by proinflammatory signals and activated by cardiac fibroblasts stretching. It signals via the calcineurin/nuclear factor of activated T-cells (NFAT) to promote myofibroblast transdifferentiation and increase collagen type I and osteopontin production (52). Osteopontin, in turn, can up-regulate the collagen cross-linking enzyme lysyl oxidase (LOX) in the heart (53), thus contributing to increased stiffness of collagen fibers.

Interestingly, in patients with limb lymphedema, high interstitial protein concentration due to chronic lymph stasis has been associated with increased numbers of fibroblasts and macrophages and with increased collagen deposition in subcutaneous tissues (54). Therefore, it is tempting to speculate that similar mechanisms may be implicated during chronic myocardial edema, contributing to the development of myocardial fibrosis.

Additionally, the cardiac lymphatic system impacts the resolution of inflammation (6,17), which in turn plays a pivotal role in aggravating myocardial interstitial fibrosis (reviewed in Schiattarella et al. [55]). Lymphatics regulate the duration and impact of immune responses through different mechanisms (reviewed in Brakenhielm and Alitalo [11]), including by actively evacuating leukocytes as well as inflammatory mediators (33). Indeed, LECs produce chemotactic gradients to attract select immune cells and express adhesion molecules essential for intralymphatic diapedesis (33,37,56). Poor lymphatic transport thus delays inflammatory resolution. Therefore, cardiac lymphatic dysfunction may also contribute to myocardial interstitial fibrosis by delaying the resolution of the immune response and perpetuating the profibrotic mechanisms triggered by activated leukocytes.

Fibrosis may conversely affect lymphatic function. Studies in secondary lymphedema have shown that chronic edema and inflammation-triggered fibrosis may further aggravate lymphatic dysfunction by directly obstructing lymphatic capillaries (57). Consequently, an antifibrotic therapy reduced lymphedema in experimental models. Additionally, fibrosis may affect flow resistance in draining lymph nodes. Interestingly, in hypertensive Dahl salt-sensitive rats, ultrastructural studies revealed an increase in collagen fibrils directly connected to cardiac lymphatics (58). Such alteration of the ECM surrounding lymphatic vessels may not only alter fluid drainage but also immune cell uptake, as LECs produce chemokines (e.g., CCL21), creating concentration gradients in the ECM necessary to guide immune cells toward lymphatic vessels (37).

Clinical Biomarkers For Lymphangiogenesis, Myocardial Edema, and Fibrosis

There are currently no validated clinical biomarkers to indicate ongoing lymphangiogenesis. VEGF-C and VEGF-D have been found to be elevated in patients with edema and/or inflammation, as well as in patients with lymph node–related cancer. Interestingly, a recent investigation in ∼2,400 patients with coronary artery disease (CAD) revealed that low circulating VEGF-C levels independently predicted all-cause mortality (59). These findings suggest that insufficient VEGF-C production may aggravate cardiac remodeling in CAD patients due to poor lymphangiogenesis. On the other hand, a small interventional study on chronic HF found worse cardiac dysfunction in patients with elevated circulating VEGF-D, but not VEGF-C, levels (60). This may be linked to the presence of pulmonary edema in severe HF, because pulmonary congestion has been associated with elevated VEGF-D plasma levels (61).

Interestingly, CMR has emerged as a reliable, reproducible, and noninvasive modality to estimate excess cardiac water and interstitial myocardial edema. In particular, T2 mapping sequences specifically detect the prolongation of the transverse component of proton relaxation time associated with water molecules (19). Indeed, a close correlation has been shown between T2 signal and cardiac water content as assessed by gravimetry (6,19). Whereas the ability of T2 to detect myocardial edema in acute ischemic as well as in inflammatory nonischemic conditions (e.g., myocarditis, takotsubo cardiomyopathy, transplant rejection) has been well established, its clinical usefulness in chronic conditions with low-grade edema has been scarcely evaluated. On the other hand, the myocardial extracellular volume (ECV), quantified using pre- and post-contrast myocardial T1 values, has been proposed as a surrogate of interstitial myocardial fibrosis (3). However, given that T1 relaxation times are prolonged by increased free water content in tissues, under conditions of edema/inflammation, the ECV may reflect a combination of both myocardial fibrosis and edema.

Importantly, CMR data can be combined with histologically proven circulating biomarkers of myocardial fibrosis (reviewed in González et al. [3]), reflecting increased deposition of collagen type I and type III fibers; the C-terminal propeptide of procollagen type I (PICP) and the N-terminal propeptide of procollagen type III (PIIINP), respectively. Interestingly, in rabbits with lymphatic obstruction, increased PICP and PIIINP levels were associated with increased interstitial myocardial fibrosis subsequent to myocardial edema (46). Finally, regarding the quality of collagen fibers, the ratio of the C-terminal telopeptide of collagen type I (CITP) to MMP-1 has been shown to reflect the degree of collagen cross-linking that increases its stiffness and resistance to degradation (3). Importantly, both CMR-derived ECV and circulating collagen biomarkers have been found to be independently associated with worse prognosis in patients with cardiac disease (3).

Therapeutic Lymphangiogenesis to Improve Lymphatic Function

We previously showed that selective stimulation of cardiac lymphangiogenesis, on the basis of targeted delivery of VEGF-C, sufficed to reduce myocardial edema, inflammation, and fibrosis post-MI in rats (Figure 3) (6). This led to improved cardiac function with partial prevention of HF development. Similarly, a recent study revealed that overexpression of adrenomedullin also improved cardiac lymphangiogenesis post-MI in mice, resulting in reduced edema and cardiac dysfunction (18). In addition, potentiation of the β1 integrin–VEGFR3 complex through conditional knockdown of ILK in mice increased cardiac lymphangiogenesis post-MI (39). Finally, a parallel strategy to improve lymphatic function and/or increase endogenous lymphangiogenesis is through anti-inflammatory therapy. This approach has recently shown promise to limit edema in patients with lymphedema (62). It remains to be determined whether the treatment improved lymphatic structure, such as remodeling of lymphatic capillary junctions, which contributes to poor lymphatic uptake under inflammatory conditions (43).

Before proposing therapeutic applications of VEGF-C in patients with chronic myocardial edema, its additional effects should be considered. VEGF-C binds 2 main receptors, VEGFR3 and VEGFR2, to induce both blood and lymphatic vascular remodeling (63). Through VEGFR3, mainly expressed by LECs, VEGF-C stimulates lymphangiogenesis, whereas through VEGFR2, mainly expressed by blood vascular endothelial cells, it stimulates angiogenesis but also vascular leakage. Indeed, therapeutic angiogenesis was the rationale of initial clinical trials with Vegfc plasmid gene delivery in CAD patients (64). More recently, an ongoing trial in CAD patients evaluates the cardiac benefit of adenoviral delivery of Vegfd with the dual aim to stimulate cardiac angiogenesis and lymphangiogenesis to improve both cardiac perfusion and lymphatic drainage (65). Of note, both VEGFR2 and VEGFR3 may be expressed by nonvascular cell types. For instance, VEGF-C acts on VEGFR3 expressed by proinflammatory macrophages to stimulate tissue recruitment (66).

Fibroblasts may also be a direct target for VEGF-C and VEGF-D. In vitro, both growth factors stimulated cardiac myofibroblasts migration and proliferation and dose-dependently increased collagen type I production (67,68). This potential profibrotic effect may be limited to isolated myofibroblasts, given that cardiac VEGF-C delivery post-MI in rats reduced interstitial myocardial fibrosis, linked to a reduction of both edema and inflammation through accelerated cardiac lymphangiogenesis (6).

Conclusions and Outlook

Cumulative evidence supports that myocardial alterations in cardiovascular diseases include both cardiomyocyte and noncardiomyocyte cell populations (fibroblasts and blood and lymphatic microvessels) and the ECM. Thus, to enhance our understanding of HF pathophysiology and spur the development of more effective therapies, novel cellular and ECM targets must be considered (69). In this conceptual framework, insufficient or maladaptive lymphangiogenesis plays a pathophysiological role in myocardial remodeling and cardiac dysfunction in ischemic and nonischemic heart disease, both in acute and chronic conditions. Impaired cardiac lymphatic function leads to myocardial edema and delayed resolution of immune responses. These alterations may, in turn, negatively affect cardiac cells. For instance, both chronic edema and inflammation can induce myocardial interstitial fibrosis (Central Illustration). Therapeutic lymphangiogenic strategies have shown promise in experimental models by reducing myocardial edema, inflammation, and fibrosis, leading to improved cardiac function post-MI. Therefore, the time has come to incorporate the evaluation of cardiac lymphatics and myocardial edema in the management of acute and chronic heart disease, in order to validate their potential prognostic relevance, as well as the therapeutic benefit of stimulating lymphangiogenesis.

Footnotes