10.1.1 Epidemiology

Myocarditis is a disease with an extreme wide range of clinical presentations, from asymptomatic to life-threatening, importantly including sudden death. This fact, in conjunction to the lack of noninvasive highly specific diagnostic methods and its overlapping symptoms with other more prevalent/incident cardiovascular diseases, makes it an underdiagnosed entity [8]. Myocarditis is the cause of sudden death in around 10% of cases [7, 13]. Important efforts had been made to develop noninvasive imaging diagnostic methods. Cardiac magnetic resonance (CMR) has been proposed as a reliable approach based on magnetic relaxation times suggesting regional myocardial edema and fluid leak. Lake Louise criteria seem to be an accurate diagnostic method based on evidence of myocardial edema on T2-weighted images plus capillary leak in T1 [4, 14–16].

Consequence of such diagnostic conundrum, retrospective studies based on autopsies have reported an extremely wide range of association of myocarditis with sudden death, from 2 to 42% [17, 18]. Importantly all of them are relatively small studies with lack of multicentric approach, thus probably biased by environmental and genetic factors.

On the other hand, prospective studies analyzing progression of myocarditis diagnosed by gold-standard biopsy parameters, an ideal clinical-epidemiologic design, report a consistent progression of myocarditis to DCM of about 30% [5, 19–21]. Post-myocarditis DCM patients have a poor prognosis, and when American Heart Association (AHA) heart failure functional classes III–IV are reached, only heart transplantation provides a definitive resolution [5].

Viral myocarditis is thought to be the most frequent type, mostly affecting children and young adults. Studies report that 60% of children diagnosed with acute myocarditis have a transplant-free survival of 10 years [21]. Also, 25–40% of DCM patients have viral genome detected in the endomyocardial biopsy [22].

10.1.2 Clinical Presentation

The presentation of acute myocarditis is extremely variable and frequently subclinical. Acute myocarditis can either lead to acute complications or progress relatively asymptomatically to chronic heart failure, then debuting clinically with chronic late-stage complications like heart failure due to DCM. As a consequence, a high level of suspicion is required by the clinician for deciding to rule out myocarditis with specific tests. The gold standard for myocarditis diagnosis is still a positive endomyocardial biopsy, an invasive procedure which requires a high clinical pretest likelihood and low risk/benefit ratio to be performed [10]. The main clinical presentation features of myocarditis are:

10.1.3 Histopathology

Myocarditis can be classified based on several parameters, mainly histopathology, time progression, and etiology. It is important to consider that once the acute inflammatory phase is surpassed, the long-term fibrosis and loss of myocardial architecture features are usually similar regardless of original type of myocarditis. The Dallas classification was defined in order to standardize the pathology reports and account for the technical issues associated with endomyocardial biopsy sampling process. It is based on conventional staining procedures (hematoxylin-eosin) and not immunohistochemistry. It categorizes myocarditis in (a) myocarditis with/without fibrosis, (b) borderline myocarditis, and (c) no myocarditis (subsequent biopsies should be classified in persistent, healing, or healed myocarditis) [25, 26]. Nevertheless, it is not particularly useful in pathophysiologic and immunologic grounds, and its usefulness has decayed over time [25]. Based on the immunopathology findings, myocarditis is classified in:

10.2.1 EAM and Eosinophilic EAM Models

Experimental autoimmune myocarditis (EAM) murine model was developed allowing to address the immunobiology of acute and chronic myocarditis. EAM has immunologic features paralleling and resembling the human entity, making it suitable to pursue pathophysiologic insights on myocarditis. Importantly, EAM mimics the most severe clinical course of giant cell myocarditis, allowing to study the entire temporal spectrum and progression of the disease to DCM. Another important feature of EAM models is a gender bias resembling the atypical pattern of human autoimmune myocarditis, in such a way that male mice are more susceptible to the induction of heart-specific autoimmunity [44].

EAM can be induced by immunization with cardiac myosin or with a myocarditogenic peptide derived from the α-cardiac myosin heavy chain emulsified in complete Freund’s adjuvant (CFA) injected twice in the first 8 days to susceptible mice strains [45, 46]. The mice strain susceptibility to EAM seems to be partially related to MHC haplotypes. The main susceptible strains are cogenic A/J background (A/J H2^a, A.BY H2^b, A.CA H2^f, and A.SW H2^s) and Balb/c (H2^d). Susceptibility of certain C57BL/10 J background strain had also been described (B10.A H2^a, B10.S H2^s, and B10.PL H2^u) [47]. In the section dedicated to viral etiology and influence of HLA haplotypes, further discussion will be made regarding the influence of MHC and non-MHC genes.

In susceptible strains, the self-peptide presentation is mainly IA restricted, altogether pointing out the importance of that specific class II chain, equivalent to human HLA-DQ. Furthermore, as will be described below, transgenic nonobese diabetic mice expressing human HLA-DQ8 instead of IA^b spontaneously develop autoimmune myocarditis.

As might be expected according to the MHC molecular biology, the H2-restricted self-peptides associated with optimal autoimmunization are strain and haplotype-specific, also demonstrating a MHC bias with potential translational implication. Thus, Balb/c mice are susceptible to immunization with α-myosin heavy chain peptide MyHCα_614–629, SWXJ to MyHCα_406–425, and MyHCα_1631–1650, whereas A/Js are susceptible to MyHCα_334–352 as well as cTnI_105–122 [48].

The rationale of the induction process is to elicit a cellular immune response simultaneously with an antigenic challenge with a cardiac-specific self-peptide in order to generate an adaptive self- and myocardium-specific immune response capable to overcome the regulatory mechanisms, thought to be suppressed in human auto-immune diseases. Myocardium-specific inflammatory process occurs peaking at day 21 after first immunization and progressing to late-stage DCM between day 40 and 60 [44, 48]. Cardiac inflammation at the peak of EAM on day 21 is characterized by a significant leukocyte infiltration in the myocardium, including innate cells like neutrophils, eosinophils, monocytes/macrophages, as well as lymphocytes, representing adaptive cellularity. The adaptive T cell-mediated response is the driving force in the development of this inflammatory process [48].

EAM models also employ the use of adjuvants and Toll-like receptor ligands. In the case of the widely studied Balb/c EAM, the protocol includes subcutaneous injection of MyHCα peptide emulsified in complete Freund’s adjuvant (CFA) and supplemented with heat-inactivated Mycobacterium tuberculosis strain H37Ra to 5 mg/mL on day 0 and 7 of the induction protocol, plus 500 ng of intraperitoneal Pertussis toxin on day 0 [44]. The requirement of coadjuvants underlines the importance of innate immunity activation, presumably via TLRs, in the generation of a robust adaptive autoimmune response [49].

We have developed recently a murine model of eosinophilic experimental autoimmune myocarditis (EoEAM) using mice deficient in IL-17A and IFnγ and IL17A−/−IFNγ−/− double knockout (DKO) mice, immunized similarly to the general EAM model [36]. EAM in IL17A−/−IFNγ−/− DKO results in a condition in which the immune response is preferentially Th2, leading to a lethal eosinophilic myocarditis.

10.2.2 Other Murine Models of Autoimmune Myocarditis

In addition to EAM, troponin-induced myocarditis was developed, in which also susceptible mice are immunized in a similar manner than described above, but using a troponin-derived peptide instead of myosin peptides. It is unclear how closely this model resembles the real pathogenic events occurring in humans. It seems that in human myocarditis the main target of autoimmunity is the myosin heavy chain. Nevertheless, the rationale of the troponin-induced model is to have the possibility to use troponin levels and anti-troponin antibodies as an alternative readout [50].

Nonobese diabetic (NOD) mice which lack expression of the murine MHC-II molecule IA (IA^b−/−) but express the human class II haplotype HLA-DQ8 develop spontaneous myocarditis. It is important to clarify that those mice are not a humanized murine model but rather a transgenic strain expressing one human HLA haplotype in substitution of the murine one [51]. Despite the xenogenic differences, the mouse antigen-presenting cells can effectively present peptides in an HLA-restricted manner to mouse T cells, then allowing certain mechanistic studies regarding HLA-biased presentation during autoimmunity. Those studies are potentially translatable to human physiology, mainly in the case of presentation of peptides derived from phylogenetically preserved self-antigens.

Murine models have also focused on the involvement of deficient PD-1 signal on T cells in the development of myocarditis [52, 53]. PD-1 is a member of the CD28 family. Its ligation with PDL1/PDL2 expressed on antigen-presenting cells during the formation of the immune synapse generates a downregulation of effector T cell activation, via anergy, apoptosis, and/or induction of regulatory properties [54]. PD-1-deficient mice (Pd1−/−) develop more severe EAM [52]. That enhanced susceptibility is associated with increased CD4 and CD8 myocardial infiltration. Similar findings were observed using the MRL-Fas^lpr/lpr, a murine model of autoimmunity with lupus-like features as consequence of a loss-of-function mutation in the Fas gene. If a PD-1 KO condition is introduced in the MLR-Fas^lpr/lpr strain, then a severe lethal myocarditis occurs spontaneously at 4–8 weeks after birth. The incidence of myocarditis was up to 96% in this murine model. The features were an increased T cell infiltration, characteristically with a Th1 bias. The latter has been proposed as a useful murine model of PD-1 influence in myocarditis development/progression [53]. This is of clinical relevance since cases of acute myocarditis have been reported in human patients as side effect of anti-PD-1 checkpoint treatment for cancer (melanoma and non-small cell carcinoma) [55–57].

10.3.1 Gender

In the case of myocarditis, the gender bias is unusual as myocarditis is more likely to occur in males, as compared with females, a feature which is mirrored in the murine models [4, 5, 8, 44, 64]. The reason for that bias is still a conundrum, but certain factors have been found as a possible explanation, importantly differences in TLR4 activation susceptibilities between sexes [4, 5, 8, 64, 65].

Using the CVB3 myocarditis murine model, it was found that myocarditis severity is significantly stronger in males, as well as the likelihood of progression to DCM [43, 66]. This occurs despite a similar viral load and replication rate among genders, suggesting the existence of sex-specific immunologic features not related to virus clearance. Males produce higher amounts of pro-inflammatory cytokines IL1β, IL18, and IFNγ during myocarditis. TLR4, an important inducer of IL1β and IL18 by monocytes/macrophages and mast cells upon ligation, is more strongly expressed in male APCs than in female hosts, in a IFNγ-independent manner. Furthermore, Tim-3, a key peripheral inducer of T regulatory cells (Tregs), is cross regulated in an inverse manner by TLR4 with estrogens as cofactor [49]. Finally, Treg development, which is proportional to Tim-3 expression/functionality, is stronger in females, even during viral myocarditis, in an estrogen-dependent manner. This fact provides not only a plausible explanation for the myocarditis gender bias but also a link between innate and adaptive immunity and the requirement for coadjuvants in EAMs, which typically involves TLR ligands [49].

10.3.2 Environment

An important factor that observationally and mechanistically has been associated with development of human autoimmunity, with some successful animal model correlates, is the environmental influence [67].

Drugs such as digoxin, cephalosporins, diuretics (like furosemide), and tricyclic antidepressants are relatively weakly associated with myocarditis, as the incidence among patient taking those drugs is low. On the other hand, anthracyclines (like doxorubicin) are strongly associated with myocardial inflammation, even in a dose-dependent manner [64, 68, 69]. The mechanisms of drug-induced myocarditis are considered consequence of pleiotropic pharmacologic properties [64, 69]. There are also case reports of myocarditis associated with drug-induced eosinophilia. Despite the fact that no extensive mechanistic studies had been performed in those cases and most of the data is based on clinical reports, it is believed that in the latter cases myocardial inflammation is a consequence of eosinophilic infiltration rather than direct drug-induced damage [70].

Physical factors like radiation are also strongly associated with myocarditis in a dose-dependent manner. The risk of actinic myocarditis increases exponentially above 2Gy [71]. Radiation induces an acute immune response mediated by TNFα and IL1β, which seems to be the initiation mechanism if radiation noxa affects heart tissues [71].

The current concept is that pleiotropic properties of those pharmacologic and physical agents induce a direct myocardial damage. Nevertheless, this direct effect is not the only cause of myocarditis but rather its trigger. After that initial damage, an innate immune response occurs associated with exposure of intracellular self-antigens and eventually modification of those proteins in the inflammatory milieu generating neo-epitopes, altogether priming a myocardial-specific autoimmune adaptive response [64, 68].

10.3.3 Viruses

Viral myocarditis is considered sometimes an independent entity. Nevertheless, it seems that once the trigger infectious noxa exerts its effect, the final effector mechanisms are similar to the ones leading to autoimmune myocarditis progression and chronic complications. Myocardium-tropic viral infection acts as a trigger and “coadjuvant” generating a sustained myocardium-specific autoimmune response. The main viruses associated with myocarditis are the CVB3, adenovirus, influenza (A, B), parvovirus B19, cytomegalovirus, and Epstein-Barr virus. All these viruses cause myocarditis with similar inflammatory features, and all could lead to DCM. The real prevalence of progression to DCM for each specific virus is not known [6, 64, 72].

The current consensus is that both immune-mediated and viral cytotoxic mechanisms play an important role. In vitro experiments have demonstrated that low-level/low-rate enterovirus replication in myocytes generates viral proteins able to produce filament disruption of myocyte structure changes resembling the features observed in vivo. This filament disruption is induced by enteroviral protease [73–75]. Thus, beyond virus cytotoxicity, a low-level viral replication can induce cardiac cell damage even in the absence of production of fully mature and infectious viral particles.

During acute infection, one of the first infiltrating leukocytes to the myocardium are Natural Killer (NK) cells, which seem to play a short-term protective role by limiting the viral infection [76]. In support of that, NK-deficient mice develop a more severe inflammatory viral myocarditis (CVB3 model) [77]. Similar to CVB3 myocarditis, NK cells are protective in EAM model since NK cell depletion led to exacerbated inflammation, fibrosis, and loss of cardiac function [78]. We have shown that the mechanism of NK cell protection is mediated by antagonizing eosinophils trafficking to the heart [78]. Eosinophils, classically considered to exert an innate response, play a final effector role in myocarditis even in parallel to a robust adaptive T cell-mediated response [36].

A “second wave” of infiltrating leukocytes during viral myocarditis is mainly comprised of T cells, peaking 7–14 days after viral infection in murine models. T cells play an important role in the clearance of viruses. Specific T cell clones have the capacity to destroy and lysate infected cardiomyocytes, according to in vitro data [72, 76]. Overall, this antiviral response has a dual effect: (a) it controls viral infection and limits its cytotoxic effects, and (b) the tissue damage associated with the T cell response leads to exposure of cryptic antigens like myosin-derived peptides, which could eventually generate an autoimmune cardiac-specific response [79].

Several facts support the involvement of autoimmunity in the progression of viral myocarditis to a chronic stage. Mice develop cardiac-specific autoantibodies in both CVB3 myocarditis and in EAM. Furthermore, the mice susceptible or resistant to chronic CVB3 myocarditis are also susceptible or resistant to EAM, respectively (see above). C57BL/10 J mice expressing H2^b haplotype can develop acute inflammatory viral myocarditis, equivalent to A/J background strains, but are protected from chronic disease and progression to DCM [62]. A similar phenomenon has been described in mice sharing MHC haplotypes but no other non-MHC antigens (A/J B10.A H2^a vs. B10.A H2^a, resistant and susceptible to virus-associated DCM, respectively) [47]. Thus, both MHC and non-MHC genes are involved in the susceptibility to autoimmune myocarditis.

10.3.4 Mimicry

Recognition of antigens by T cell receptors (TCR), B cell receptors (BCR), and antibodies relies on the existence of specific complementarity of molecular regions of proteins, called epitopes, and the antigen-binding regions of the enounced receptors and antibodies (in a MHC-restricted manner in the case of TCR). Mimicry implies molecular similarity between non-self (microbial or alloepitopes) and self-epitopes, and the consequent cross-reactivity with T cell clones and/or antibodies (BCR are antibody-like surface receptors) [80, 81]. A paradigmatic example is the association of infections of Campylobacter jejuni with the development of Guillain-Barré syndrome [82].

It is still unclear if the molecular mimicry phenomenon is critical in the pathogenic process of myocarditis, but some studies suggest at least a partial contribution. The main proposed target of autoimmune response in myocarditis is the heavy chain of the myosin, specifically the isoform expressed in myocytes αMyHC [83, 84]. Multiple myosin isotypes had been identified, but myosin sequence is relatively phylogenetically preserved [85]. Using an A/J background murine model (H2^k haplotype), it was found that autoreactive T cell clones recognizing αMyHC peptides (specifically 334–352) in an IA^k-restricted manner have cross-reactivity with microbial epitopes derived from Bacillus spp., Magnetospirillum gryphiswaldense, Zea mays, and, even more important for its human clinical implications, Cryptococcus neoformans [84]. The latter fungi had shown a breakout within the last decades in association with the AIDS pandemic. That might have implication on the pathogenesis of myocarditis, mostly in patients susceptible to C. neoformans infections under chemotherapy or carrying clinical HIV infection.

Another particular kind of molecular mimicry was described in the context of myocarditis as similarity between self-antigens, myosin epitopes, and β-adrenergic receptors. Therefore, the autoantibodies generated during myocarditis against myosin peptides may cross-react with adrenergic receptors and exert an activating adrenergic effect upon ligation, thus potentially contributing to chronic sympathetic-mediated cardiac damage [86].

10.3.5 Exposure of Encrypted Self-Antigens

Another concept demonstrated to play a role in the development of autoimmunity is the exposure of self-antigens which are encrypted and unavailable to the immune system under physiologic conditions [87, 88]. Cardiac myosin (specifically αMyHC) is one of the most important self-targets in myocarditis. Importantly, myosin, which belongs to the contractile apparatus of the myocytes, is not significantly exported to the extracellular matrix.

Myh6, the gene encoding αMyHC, has been shown to not be expressed in medullary thymic epithelial cells (mTECs), which are the main responsible for T cell thymic negative selection, via promiscuous self-antigen presentation. αMyHC is not expressed neither in peripheral lymphoid stromal cells. As consequence, in physiologic conditions, specific self-reactive anti-αMyHC T cell clones escape from negative selection, reaching periphery as autoreactive cardiac-specific clones. That is an important feature predisposing development of myocarditis in mice and humans as long as other events or risk factors take place leading to release of αMyHC, a cryptic self-antigen. In fact, it was shown that transgenic mice, expressing Myh6 in mTECs, are protected from EAM induction in association with clonal deletion of those autoreactive clones [89].

Either myocardial infection, toxins, ischemia or other insults can lead to myocardial damage and subsequent exposure of intracellular proteins (cryptic epitopes). That exposure, in the context of a proper pro-inflammatory microenvironment, triggers the adaptive immune response leading to myocarditis in genetically susceptible individuals [6, 72, 76]. Thus, different triggers could lead to myocarditis and eventually chronic autoimmune myocarditis.

10.3.6 MHC and Non-MHC Bias

Human Leukocyte Antigen (HLA) genes (human form of Major Histocompatibility Complex, MHC) are one of the most polymorphic genes in humans, as well as the coresponding MHC genes are in most mammals. This genetic variability has functional consequences in terms of the affinity of MHC molecules for specific peptides. Despite MHC molecules being promiscuous in terms of peptide presentation, certain biases exist as consequence of the avidity and affinity of MHC molecules for specific peptides [2].

The association of certain autoimmune diseases with specific HLA haplotypes is considered to be consequence of a higher affinity of certain haplotypes for preserved protein self-products.

HLA haplotypes are not sufficient, and may not be main determinants, in myocarditis initiation and progression. Nevertheless, certain associations between DCM of any cause and HLA haplotypes had been found. Specifically, HLA-DR4 is statistically associated with DCM based on retrospective studies [90–92]. Other DCM-HLA-positive associations had been reported, mainly HLA-DR12, DR15, and DRB*0601, as well as negative associations (HLA-DR11, DQB1*0301) [93–95].

Experimental facts observed in murine models of EAM and CVB3 myocarditis also support the influence of MHC haplotypes in the development of myocarditis and its progression to DCM. Mice with b haplotypes in A/J background are less susceptible than other non-H2^b A/J strains to EAM and CVB3 myocarditis including the acute inflammatory phase and DCM [96]. Furthermore, that b-associated protection on A/J background parallels development of lower titers of cardiac-specific autoantibodies [62, 96].

Similarly, transgenic nonobese diabetic mouse strain expressing human HLA-DQ8 instead of IA^b develops spontaneous myocarditis which progresses to DCM even with electrophysiologic disturbances like heart block (see Sect. 10.2) [51, 97]. Nevertheless, it is of notice that HLA-DQ8 is not within the haplotypes associated with DCM in human studies, which might limit the translational findings of this murine model and/or underscore the importance of other non-MHC genes.

There is a lack of myocarditis-specific association of MHC haplotypes in certain murine models of myocarditis as was shown by using specific strains sharing MHC haplotypes (MHC-full matched), but differing in non-MHC background-associated genes has been performed. A.SW (A/J background) and B10.S (B10 background), both H2^s, have differences in myosin-induced EAM susceptibility, being B10.S protected as compared with susceptible A.SW [98]. Similarly, mice with b haplotypes differ in susceptibility to EAM. C57BL/10 J are resistant, while A/J are very susceptible to EAM. That strongly suggests the influence of other non-MHC features in the development of myocardial-specific autoimmunity. Using simple sequence length polymorphism (SSLP) markers in the murine genome, two non-MHC loci were found to be associated with the development of EAM in H2^s equivalent strains. Those genes, named in mice Eam1 and Eam2 (located in proximal chromosome 1 and distal chromosome 6, respectively), are linked to myocarditis development. Interestingly, those loci and their human equivalents had been associated with SLE and diabetes, as well as autoimmune experimental encephalitis and orchitis in mouse [98]. Other studies using the CVB3 murine model and also SSLP tracking system found other non-MHC loci associated with viral myocarditis: Vms1 (chromosome 1), Vms2 (chromosome 4), and Vms3 (chromosome 3) [99].

Finally, a specific 14 bp deletion in a region of the HLA-G gene (a human non-classical MHC I molecule) was reported to be associated with DCM. Nevertheless, as linkage disequilibrium exists between HLA-G and HLA-DR and DQ and considering that the reported HLA-G 14 bp deletion occurs in the 3′-untranslated region of the gene, it remains unclear if the HLA-G bias corresponds to a real mechanistic association, or just a correlation whose mechanism relies on class II-linked haplotypes [100, 101].

Other polymorphisms have been identified in association with autoimmunity, mainly on the CTLA-4, PD-1, and ICOS genes. All these genes encode for functional proteins involved in the regulation of T cell activation mainly via induction of apoptosis and/or anergy. Specifically, mutations in PD-1 and ICOS are associated with the development of myocarditis in murine models but in the context of broader systemic autoimmune features [52, 53, 102, 103]. The case of PD-1 deserves particular attention considering the outbreak of the checkpoint blockade in cancer therapy. PD-1 blockade has been shown to provide benefit in certain cancers like non-small cell lung carcinoma and melanoma [55, 56]. Myocarditis has been found to be an infrequent side effect of PD-1 blockade. The histological features were that of acute lymphocytic myocarditis [57]. This validates the importance of the disruption of the PD1/PDL1 axis in the development of myocarditis observed in mice.

10.3.7 Autoimmune Regulator (AIRE)

Autoimmune regulator (AIRE) allows a promiscuous expression of self-antigens by medullary thymic epithelial cells (mTECs) [104, 105]. AIRE is involved in central deletion of effector T cells (Teff) and thymic Treg (tTreg) induction but also in peripheral Teff anergization and pTreg differentiation [104–106].

The absence of AIRE expression in murine models (AIRE−/− mice) leads to an increase of autoreactive effector T cells in the periphery [107]. AIRE also influence self-tolerance by directing autoreactive bone marrow-derived T cell clones into regulatory functions [107].

Taking these concepts to the field of myocardial autoimmunity, some experimental data have demonstrated that defective central (thymicdependent) tolerance and increase of autoreactive Teff (specific for myosin-derived epitopes) as well as decreased thymic Tregs (tTregs, formerly nTregs) with the same specificity are involved in the inflammatory phase of EAM [106, 108]. HLA-DQ8+ IA^b−/− NOD mice develop spontaneous myocarditis with cellular and humoral autoimmune responses directed toward epitopes of the α isoform of myosin heavy chain (αMyHC). The same model showed that EAM is dependent on the T cell-mediated anti-MyHC response [109]. Anti-myosin-specific T cells represent the majority of the myocardium-infiltrating Teff [89]. The development of myocarditis mediated by MyHC-specific Teff was associated with lack of expression of αMyHC by mTECs. With more complex genetic manipulations of that murine strain, it was shown that susceptibility to EAM was abrogated by expression of MyHC by mTECs, paralleling a decrease in that specific Teff autoreactive clones in the periphery. Further studies about the functional impact of AIRE in the expression of myosin by human mTECs as a factor determining development of autoimmune myocarditis are needed [89].

10.4.1 Adaptive Cellular Response

The adaptive cellular immune response is characterized by its antigen specificity and high inflammatory efficiency. Its major mediators are the T cells, which specificity relies in the molecular structure of clonal TCRs, able to “recognize” characteristic peptide/MHC complexes.

The classic dichotomist Th1/Th2 model does not provide an accurate representation of the normal immune response [111, 112]. That became evident in the last decades after discovery of other important pro-inflammatory Teff subsets, like Th17 (but also including Th9 and Th22), as well as anti-inflammatory subpopulations, mainly Tregs and Tr1. Furthermore, a high plasticity of T cell subsets has been described, involving not only central (thymic) but also peripheral T cell fate induction [112]. Dysfunction of Th17 and Treg has been widely found to be a key factor in autoimmunity, including myocarditis. Importantly Treg and Th17 activation/differentiation shares certain cytokine requirements, remarkably TGFβ, needed for both subsets but in a concentration-dependent manner. In the presence of TGFβ, IL6/IL-1β vs IL10 balance is responsible for the Th17 vs Treg differentiation [112].

10.4.1.3 Th17 Response

IL17A, the main pro-inflammatory Th17-derived mediator, unexpectedly is not necessary for development of acute inflammatory stages of myocarditis in EAM model [125]. On the other hand, we have found that IL-17A is strictly required for progression to DCM, cardiac fibrosis, and loss of cardiac function. In other words, the acute inflammation can take place in the absence of IL17A; however without IL17A, myocarditis does not progress to DCM [125].

Importantly, the Th17 response required for progression from myocarditis to DCM is not only dependent on leukocytes. We demonstrated that beyond inflammatory Ly6C^high monocytes, also cardiac fibroblasts play an important role in the inflammatory process. IL17A−/− mice, protected from DCM but not acute EAM, have a diminished myocardial infiltration of neutrophils and Ly6C^high inflammatory monocytes [126]. Interestingly, a conversion of Ly6C^high monocytes to Ly6C^lo monocytes protects WT host from DCM, strongly suggesting that Ly6C^high monocytes are associated with the pro-fibrotic IL17A effect. We have found that granulocyte-monocyte colony-stimulating factor (GM-CSF) is required, in conjunction with IL17A, for Ly6C^high monocyte infiltration in myocardium during EAM. We showed that IL-17A is able to induce high amount of GM-CSF and other myeloid cytokines and chemokines from cardiac fibroblasts. Thus, we have discovered an active Th17-fibroblast-monocyte cross talk in the pathogenesis of myocarditis. To summarize, T cell-derived IL17A plus fibroblast-derived GM-CSF contributes to Ly6C^high monocyte chemotaxis/activation, which plays an active role in fibrosis and sustains inflammatory process (Fig. 10.2) [126].

Open in a separate window

Fig. 10.2

Schematic immunopathogenesis of post-myocarditis DCM development. CD4 T cell autoimmune response induces a Th17-dependent pro-inflammatory process associated with DCM development. T cell-derived IL17A induces cardiac fibroblasts to produce GM-CSF, which in turns activates monocytes toward a highly inflammatory function. Those activated monocytes are required final effectors in the progression to chronic tissue damage and DCM

As described above, a very specific milieu of cytokines is needed for Th17 cell differentiation and establishment of the so-called Th17 environment. That milieu includes IL6, IL23, and TGFβ. Importantly, IL23 is needed for Th17 terminal differentiation and Th17 “stabilization” [111, 112]. We demonstrated the requirement of IL23 (p19) during EAM development [117]. Specifically, CD4 T cell stimulation by IL23 is required during the acute phase of autoimmunity for an effective myocarditis (EAM) and consequent DCM development. The key effect of such IL23 stimulation was proven to be the GM-CSF production by T cells. Furthermore, IL23 effect is only transiently required during the acute phase (early after EAM induction), as later influence cannot restore the EAM development [117].

An important translational study supporting the latter concepts was published recently [127]. That study supports the influence of Th17 in human myocarditis and its progression to DCM. Th17 CD4 T cells were found to be associated with myocarditis, influencing inflammation via IL6, TGFβ, and IL23. Also, an association exists with GM-CSF producing monocytes (CD14+). This interaction was related with TLR2 expression on those inflammatory monocytes, an interestingly finding unveiling the role of this particular TLR type in human myocarditis. Consistently, a decreased classic Treg population (Foxp3+) was observed in association with the enounced features [127].

Those seminal studies made evident potential Th17 response-focused therapeutic targets, not only to treat acute inflammatory myocarditis but also its late-stage complications.

10.4.2 Adaptive Humoral Response

Despite the existence of clearly humoral-mediated autoimmune diseases (like post-streptococcal glomerulonephritis, Berger disease, and Goodpasture disease), the influence of those autoantibodies on the pathogenesis of most autoimmune diseases is still controversial, at least on its initiation phase. Myocarditis is a cellular-mediated process, and several myocarditis-associated autoantibodies are specific to encrypted antigens that are not available to antibody ligation without tissue damage and cell disruption [87]. As consequence, the generation of autoantibodies by clonal selection of self-reactive B cell clones is secondary to T cell activation, T-B cross talk, and release of intracellular or matrix-encrypted proteins after tissue damage [131, 132].

In myocarditis, this lack of pivotal role during initiation is supported by the fact that up to two-thirds of acute myocarditis patients have no detectable cardiac-specific antibodies at the moment of diagnosis or clinical onset [83, 132]. On the other hand, its correlation with disease progression and detection in symptom-free relatives of DCM patients suggest that they occasionally may be early markers paralleling myocardial damage after its initiation, with minimal to null causative influence in the initial phase [132].

Eighty percent of patients with DCM-related myocarditis develop cardiac-specific antibodies, but also up to 60% of patients with heart failure of any cause have circulating antibodies targeting heart-specific epitopes. The latter supports the idea of the necessity of tissue damage of any cause for development of humoral autoimmunity, as well as the paucity of its inflammatory influence [83, 132–134].

A plethora of autoantibodies has been described in myocarditis, including anti-αMyHC, troponin I and T, and β1-adrenergic receptor. Also, non-cardiac-specific autoantibodies are frequently found in myocarditis patients, including anti-actin, tropomyosin, laminin, and anti-muscarinic receptor [4]. Of those, the ones receiving most attention are the anti-αMyHC and the anti-adrenergic receptor. Anti-αMyHC antibodies have been described to correlate with disease progression (including diminished titers after adequate clinical response) and are also found in murine models [132]. Anti-adrenergic receptor antibodies are also correlated with disease progression, and interestingly a significant proportion of the clones are activating antibodies, so those antibodies might contribute to hemodynamic alterations via autonomic modulation [4, 83, 132, 133, 135, 137].

Antibodies targeting specific receptors might induce persistent activation of such receptors, i.e., induce a sustained positive intracellular signaling, similar to the one generated by physiologic ligands, but abnormally sustained over time. This is not the case of all receptor-specific antibodies but seems to be the case of several clones involved in myocarditis targeting adrenergic receptors [133]. The case of cross-reaction of anti-myosin antibodies with adrenergic receptors was already described above in the section dedicated to mimicry [84].

In the case of activating β1-adrenergic antibodies (with a prevalence of 70–80% among patients with dilated cardiomyopathy of any cause and 60% on patients with myocarditis), they seem to exert their persistent adrenergic effect by stabilizing an activated molecular conformation of the receptor upon ligation. Also, the cross-link of two receptors, generating a stable dimerization, contributes to that effect [132, 133].

A pathogenic sustained β-adrenergic activation has been demonstrated to be able to induce cardiomyocyte apoptosis [137, 138]. Furthermore, once systolic dysfunction begins, the increased adrenergic tone (sympathetic) has deleterious hemodynamic effects [139, 140]. It is believed that beneficial effect of β blocker drugs in patients with heart failure of any cause is related both with the improvement of hemodynamic status and the decrease of the proapoptotic effect. Nevertheless, theoretically, such drugs cannot reverse per se the stabilization of the receptors induced by the activating autoantibodies. In that regard, the turnover rate and upregulation of those receptors must be considered [140].

Therapeutic approaches are focused on the depletion or neutralization of anti-adrenergic activating autoantibodies. Plasmapheresis and immunoadsorption of those antibodies were tested in case-control clinical studies in patients with dilated cardiomyopathy [141]. Interestingly, also benefit from the removal of cardio-depressant autoantibodies by immunoadsorption was also found [142]. Patients receiving immunoadsorption have improved cardiac function. After a 12-month follow-up, anti-β-adrenergic receptor autoantibodies did not return to the original titers [143].

Another treatment strategy with a similar goal is the neutralization of those antibodies using aptamers in the apheresis technology. As aptamers are synthetic oligonucleotide ligands with high specificity to targets like the Fab region of antibodies, its use is proposed as a mean to optimize the removal of autoantibodies by apheresis and decrease toxic and immunogenic complications described for immunoadsorption [144, 145]. The overall utility of this approach has not been fully elucidated.

10.5.1 Immunosuppression

The rationale of this therapeutic approach is to target the pro-inflammatory mediators of the disease. Immunosuppression is used in treatment of myocarditis types considered to be autoimmune such as giant cell myocarditis. So far, broad immunosuppressive drugs rather than targeted treatments have been used. The main drugs tested for myocarditis are steroids (prednisone, prednisolone), cyclosporine, and azathioprine [9, 30]. These drugs are mostly used in combination; however, the benefits for individual subtypes of myocarditis need to be evaluated in proper clinical comparative studies. Some studies showed significant or at least some improvement in terms of ejection fraction [155], but another study reported no benefit on ejection fraction or survival as compared with placebo [156]. Notwithstanding, differences in the study design exist, and importantly the study reporting benefit was made on nonviral myocarditis cases, whereas in the one reporting no benefit no viral status discrimination was made.

Another interesting study analyzed the effect of muromonab-CD3 (a monoclonal antibody targeting CD3) plus cyclosporine and steroids in a prospective cohort of 12 patients. This study included specifically patients with giant cell myocarditis. Despite there was no control group (which is understandable due to the high severity and lethality of this type of myocarditis), an improved survival was reported [28–30, 157].

10.5.2 Immunomodulation

As described in the section of humoral response, plasmapheresis/immunoadsorption has been studied by several groups. A consistent benefit was observed in hemodynamic terms, mainly improvement of ejection fraction. Those protocols include the coadministration of immunoglobulin. Interestingly, those studies enrolled patients with DCM, which supports the hemodynamic deleterious effect of cardiac autoantibodies. Importantly, DCM patients of any cause were included, which expands the usefulness of this approach to non-myocarditis cases [141, 159–161].

High doses of immunoglobulin exert a beneficial effect (improved survival and ejection fraction) in acute myocarditis patients. Several data come from small case studies [153], but a randomized controlled study on acute myocarditis patients supports the observation [161]. It is of notice that one controlled study reports conflicting data by not finding improvements of the ejection fraction [162]. Also, in those studies, no clear etiologic definition was made.

10.5.3 Antiviral

The main antiviral drugs studied in myocarditis have been acyclovir, peramivir, ganciclovir, ribavirin, and artesunate, in well-defined viral myocarditis patients (etiology was specifically determined: parvovirus B19, influenza, cytomegalovirus, parainfluenza, herpesvirus, respectively) [6, 72, 153]. All of them have been either small studies or case reports, but overall the benefit of antiviral treatment seems to be consistent in terms of ejection fraction, survival, and decreased viral load. All protocols had included immunosuppressive and/or immunomodulatory treatment, as steroids and/or immunoglobulin [153]. The latter is consistent with the described importance of the immune-mediated damage in conjunction with viral cytotoxic effects in the development of viral myocarditis.

Also, interferon α and β were analyzed in viral myocarditis. Interferon α was found beneficial in enterovirus-confirmed DCM, based on hemodynamic parameters and viral clearance. Interferon β significantly improved ejection fraction and viral load in a cohort of acute viral myocarditis [163], whereas no benefit was observed in a cohort of patients treated at DCM stage [164].