Pathogenesis of inflammatory myopathies

INTRODUCTION — The inflammatory myopathies are a group of disorders sharing the common feature of immune-mediated muscle injury. Clinical and histopathologic distinctions between these conditions suggest that different pathogenic processes underlie each of the inflammatory myopathies. The most common of these disorders include:

●Dermatomyositis (DM)

●Polymyositis (PM)

●Antisynthetase syndrome

●Immune-mediated necrotizing myopathy (IMNM)

●Inclusion body myositis (IBM)

●Nonspecific myositis

Other subtypes of inflammatory myopathy include eosinophilic myositis and granulomatous myositis. Many patients with inflammatory myopathies cannot be assigned to any category and may be classified as having nonspecific myositis [1,2]. The precise mechanisms leading to tissue injury in the inflammatory myopathies are incompletely defined in most cases.

The pathogenesis of the inflammatory myopathies, including DM, IBM, IMNM, antisynthetase syndrome, and PM, is reviewed here. The clinical manifestations, diagnosis, and treatment of the inflammatory myopathies, the pathogenesis of juvenile DM and PM, and overlap syndromes associated with systemic rheumatic diseases are discussed elsewhere. (See "Overview of and approach to the idiopathic inflammatory myopathies" and "Clinical manifestations of dermatomyositis and polymyositis in adults" and "Initial treatment of dermatomyositis and polymyositis in adults" and "Treatment of recurrent and resistant dermatomyositis and polymyositis in adults" and "Malignancy in dermatomyositis and polymyositis" and "Juvenile dermatomyositis and other idiopathic inflammatory myopathies: Diagnosis" and "Juvenile dermatomyositis and other idiopathic inflammatory myopathies: Epidemiology, pathogenesis, and clinical manifestations" and "Juvenile dermatomyositis and polymyositis: Treatment, complications, and prognosis" and "Clinical manifestations and diagnosis of inclusion body myositis" and "Management of inclusion body myositis" and "Undifferentiated systemic rheumatic (connective tissue) diseases and overlap syndromes".)

PATHOGENESIS OF MAJOR SUBSETS OF MYOPATHIES — The development of genomic technologies spawned by the Human Genome Project have opened a new era in our understanding of the pathogenesis of inflammatory myopathies. These technologies include transcriptomic profiling by microarray and ribonucleic acid (RNA) sequencing (RNA-seq), protein profiling by proteomics and refinements in flow cytometry such as cytometry by time of flight (CyTOF), and hybrid RNA/protein methods such as cellular indexing of transcriptomes and epitopes by sequencing (CITE-seq). These technologies have enabled deep, unbiased discovery of pathophysiologic mechanisms in inflammatory myopathy muscle tissue. The linking of these mechanisms to clinical features, blood biomarkers, and conventional microscopy is radically changing previous views of the pathogenesis of inflammatory myopathies. The pathogenetic mechanisms are discussed further below.

Dermatomyositis — The precise pathogenic mechanisms responsible for dermatomyositis (DM) are unknown. The application of genomic technologies to DM skin and muscle samples has suggested that DM pathology results from exposure to a type 1 interferon (IFN) causing capillary, myofiber, and keratinocyte injury [3-7].

By contrast, an earlier model of DM myofiber injury resulting from an antibody- and complement-mediated microangiopathy had predominated, largely on the basis of microscopic observations of muscle biopsies [8,9]. The microscopic pathology of DM is unique among all muscle diseases, consisting of perifascicular atrophy, capillary abnormalities and loss, and perimysial abnormalities. Perifascicular myofibers, located at the boundaries between fascicles and perimysial connective tissue, are preferentially affected relative to central fascicular myofibers. These fibers are atrophic and, by microscopic histochemical techniques, express molecules typically associated with "regeneration," such as neonatal myosin (MHY8), vimentin, and natural cell adhesion molecule (NCAM), though it is unlikely these myofibers are regenerating.

Type 1 interferon — There is substantial evidence that the induction of type 1 IFN-inducible gene products, both transcripts and proteins, driven by IFN-beta (IFNB), is highly and specifically associated with DM [3,4,6]. The human type 1 IFNs consist of IFN-alpha (IFNA), IFNB, IFN-kappa (IFNK), and several other subtypes, but do not include IFN-gamma (IFNG).

●Type 1 IFN-inducible transcripts in muscle – Large-scale profiling, using microarrays, of messenger RNA (mRNA) transcripts abundant in inflammatory myopathy muscle has shown very high levels of both immunoglobulin transcripts and many transcripts that are known to be induced by type 1 IFN, IFNA, and IFNB [7,10-12]. Type 1 IFN-inducible transcripts are the single most upregulated pathway present in DM muscle, with approximately 90 percent of the highest differentially upregulated transcripts belonging within this class. The degree of upregulation of most of these transcripts is unique to DM among inflammatory myopathies [6]. For example, the transcript for IFN-stimulated gene 15 (ISG15) was 570-fold higher, and myxovirus resistance protein A (MxA) was 280-fold higher, in DM than in normal muscle, compared with approximately 5-fold higher levels compared with normal for each in inclusion body myositis (IBM) [13].

●Type 1 IFN-inducible proteins in muscle – Among patients with DM, the type 1 IFN-inducible protein, MxA, has been observed to be densely expressed on perifascicular myofibers and sometimes on all myofibers (picture 1) [7]. MxA expression in myofibers is specific to DM, and MxA is also overexpressed in the epidermis [14]. The most sensitive and specific diagnostic histologic marker of DM is the expression of MxA, a type 1 IFN-inducible protein, by perifascicular myofibers [7,15,16].

●Type 1 IFNs in muscle and blood – The high specificity of the presence of type 1 IFN-inducible transcripts and proteins in DM muscle suggests that DM muscle is exposed to high concentrations of a type 1 IFN. Which specific type 1 IFN(s) drive this induction in DM muscle is unknown and will likely only be determined by clinical trials of selective type 1 IFN neutralization. The available biomarker evidence strongly suggests that IFNB is the driving cytokine. Blood measurements of type 1 IFN protein serum levels have shown association of IFNB but not IFNA with DM type 1 IFN signature [17] and disease activity [18]. Most remarkably, RNA-seq muscle transcript profiling has shown remarkable specificity for IFNB muscle transcript in DM, among all type 1 IFN subtypes and all categories of inflammatory myopathy (figure 1). The cellular source of IFNB and potentially other type 1 IFNs in DM muscle is unknown, but is likely fibroblasts in the muscle perimysial region or skin dermis, or myofibers and keratinocytes themselves. An alternative possibility is from locally infiltrating plasmacytoid dendritic cells (pDCs) [7,19].

Relying upon these and other observations, a model of the pathogenesis of DM resulting from the induction of type 1 IFN-inducible gene products after exposure to type 1 IFNs can be constructed [5]. Production of a type 1 IFN, particularly IFNB by fibroblasts, myocytes, and keratinocytes, leads to sustained accumulation of intracellular (within myofibers and keratinocytes) type 1 IFN-inducible transcripts and proteins. These injure myofibers and keratinocytes through mechanisms unknown.

Autoantibodies — Autoantibodies with affinities for various antigens, such as the aminoacyl transfer RNA (tRNA) synthetases, have been reported in DM and other inflammatory myopathies [20-22]. The pathogenic role of any of these autoantibodies is uncertain and continues to be explored, although the detection of these autoantibodies may be useful clinically because of their association with particular clinical syndromes. (See "Overview of and approach to the idiopathic inflammatory myopathies", section on 'Myositis-specific autoantibodies'.)

DM has increasingly been recognized as constituting a number of subtypes defined by the presence of specific autoantibodies [23]. The earliest identified DM syndrome-associated autoantibodies were the antisynthetase syndromes, typified by anti-Jo-1, though these are associated with polymyositis (PM) as well [20]. More recent DM subtypes associated with anti-melanoma differentiation-associated protein 5 (anti-MDA5), anti-transcription intermediary factor 1-gamma (anti-TIF1-gamma), anti-nuclear matrix protein 2 (anti-NXP2), and anti-small ubiquitin-like modifier activating enzyme (anti-SAE) have been defined with specific clinical manifestations. The role of these autoantibodies in DM pathogenesis is unknown.

Polymyositis — The definition of PM has evolved since the disease was first described, resulting in challenges identifying true pathogenetic mechanisms. PM was first reported in 1886. The recognition of skin rash in some patients led to distinction of DM by 1891, but PM remained the sole diagnostic category for patients with myositis other than DM for the next 90 years. Subsequently, there has been a progressive splitting off of syndromes from PM into more specific categories, first with IBM (1978), then with immune-mediated necrotizing myopathy (IMNM; 1991), and with antisynthetase syndrome. What remains of this initially highly heterogeneous category of PM is a matter of differing opinions. Some investigators prefer the term nonspecific myositis [1] or unclassified myositis [24] over PM to reflect its continued heterogenous nature. In addition, some patients with muscular dystrophies related to dysferlin and calpain-3 mutations have been misdiagnosed with PM because of the presence of immune cells in muscle biopsies. (See "Clinical manifestations of dermatomyositis and polymyositis in adults" and 'Immune-mediated necrotizing myopathy' below.)

The heterogenous nature of a lumped category like PM has created particular confusion with regard to the pathogenesis of the diseases included in it. Numerous publications have viewed PM as a disease in which lymphocytes, particularly CD8+ cytotoxic T cells, invade myofibers, with similar or identical immunologic mechanisms as are seen in IBM. Unfortunately, the misdiagnosis of patients with IBM as having PM largely contributed to this view. Some investigators have reported that almost all patients with initial lymphocytic invasion of myofibers have IBM after follow-up [24].

Antisynthetase syndrome — As with IMNM, antisynthetase syndrome has historically been lumped within the category of "polymyositis" (PM), but its distinctive features warrant separate classification [25] (see "Overview of and approach to the idiopathic inflammatory myopathies", section on 'Antisynthetase syndrome'). This disease is characterized by multiple-organ involvement in addition to myositis (and sometimes without myositis), primarily interstitial lung disease, non-erosive arthritis, Raynaud phenomenon, and "mechanic's hands." The last feature has variably been interpreted as a feature of DM. Autoantibodies directed to eight different aminoacyl tRNA synthetases define this category, with anti-Jo-1 being the most common. Multiple immune components, including T cells, have been suggested as pathogenic factors [26] (see "Overview of and approach to the idiopathic inflammatory myopathies", section on 'Myositis-specific autoantibodies'). The role of autoantibodies is unclear, as mouse models of myositis induced by immunization with histidyl tRNA synthetase are not dependent on the development of antibody responses [27] and have been hypothesized to be mediated by innate immune mechanisms or by the action of histidyl tRNA synthetase as a chemokine [27]. More generally, activation of multiple components of the innate and adaptive immune system may play a role in the pathogenesis of antisynthetase syndromes [28].

Immune-mediated necrotizing myopathy — IMNM was identified as distinct from "polymyositis" (PM) in the early 1990s [29]. Three subtypes of IMNMs are recognized [30]:

●Anti-signal recognition particle (anti-SRP) autoantibody IMNM [31]

●Anti-3-hydroxy-3-methylglutaryl-coenzyme A reductase (anti-HMGCR) autoantibody IMNM [32]

●Anti-HMGCR/SRP-negative IMNM

Histologically, IMNM is characterized by scattered necrotic and regenerating fibers (often many more regenerating fibers than fibers undergoing necrosis) with relative absence of lymphocyte infiltration and abundance of macrophage infiltration [33]. Anti-HMGCR myopathy is strongly associated with both the class II major histocompatability complex (MHC) allele DRB1*11:01 and exposure to statin drugs, suggesting it results when genetically susceptible individuals upregulate HMGCR, particularly after statin exposure [34]. The pathophysiology of anti-SRP autoantibody IMNM is less understood. The mechanism of myofiber injury for all forms of IMNM is even less clear, speculatively related to autoantibody binding to the surface of myofibers and subsequent complement-mediated cellular injury [34]. Notably, in anti-HMGCR, anti-SRP, and in those IMNM cases with no serum antibodies, muscle biopsies reveal expression of MHC class 1 (MHC1) and membrane attack complex (MAC) deposition on the sarcolemma of non-necrotic fibers, suggesting that muscle destruction might be complement mediated. Deposition of MHC1 and MAC on sarcolemma of non-necrotic muscle fibers is not seen in toxic myopathies nor other autoimmune inflammatory myopathies. (See "Statin muscle-related adverse events".)

Other myopathies — Other forms of myositis have been described, some of which exhibit characteristic histologic features. These include:

●Drug-induced myopathy – Several drugs appear to contribute to the development of inflammatory myopathy, including statins, IFNA and IFNB, tumor necrosis factor (TNF) inhibitors, L-tryptophan, and others [35-48]. Interpreting causality is often confounded by the presence of the underlying disease for which the drugs are being used. This topic is discussed in detail elsewhere. (See "Drug-induced myopathies" and "Drug-induced myopathies", section on 'Pathogenesis' and "Statin muscle-related adverse events" and 'Immune-mediated necrotizing myopathy' above and "Risk factors for and possible causes of systemic sclerosis (scleroderma)", section on 'L-tryptophan'.)

●Eosinophilic myositis – Eosinophilic myositis is sometimes classified as a form of PM. The only distinctive feature is the presence of eosinophils in muscle biopsy sections and, occasionally, blood eosinophilia. Chronic parasitic infection is one known cause of this syndrome. Remarkably, some patients with eosinophilic myositis have mutations in calpain-3 [49,50]. Mutations in this gene also result in an inherited muscular dystrophy syndrome (limb-girdle muscular dystrophy type 2A). The mechanisms by which calpain-3 mutations lead to recruitment of eosinophils to muscle are unknown.

●Granulomatous myositis – Granulomatous myositis is a clinical syndrome of distal asymmetric weakness and the presence of granulomas in muscle biopsy sections (picture 2). It may occur with systemic involvement leading to a diagnosis of sarcoidosis [51] or graft-versus-host disease [52,53], or it can be idiopathic, without another associated illness [51,54]. Macrophages, lymphocytes, and plasma cells are present in the affected muscle, but the mechanisms of idiopathic granulomatous myositis are unknown [54]. Case series have suggested as many as half of patients who have granulomas on muscle biopsy actually have IBM (both clinically and by repeat biopsy) [55,56].

●Myositis associated with graft-versus-host disease – Graft-versus-host disease may be associated with myositis, characterized pathologically as a number of syndromes including DM (picture 3) [57-60], PM [61], and granulomatous myositis [52,53]. The pathogenesis of these syndromes is poorly understood, but the presence of infarction in some of these clearly indicates vascular injury (see panels C and D in the picture) (picture 3).

SUMMARY

●The inflammatory myopathies, including dermatomyositis (DM), polymyositis (PM), antisynthetase syndrome, immune-mediated necrotizing myopathy (IMNM), inclusion body myositis (IBM), and other myopathies and related disorders, share the common feature of immune-mediated muscle injury. Clinical and histopathologic distinctions between these conditions suggest that different pathogenic processes underlie each of the inflammatory myopathies, but the precise mechanisms leading to tissue injury in the inflammatory myopathies are incompletely defined. (See 'Introduction' above.)

●The precise mechanisms responsible for DM are unknown. The marked type 1 interferon (IFN) molecular abnormalities present strongly suggest a pathogenesis including overproduction of a type 1 IFN, with the available evidence suggesting IFN-beta (IFNB) predominantly. The pathogenic role of myositis-specific autoantibodies is uncertain. (See 'Dermatomyositis' above and 'Type 1 interferon' above and 'Autoantibodies' above.)

●The heterogeneous nature of a category like PM has created particular confusion with regard to the pathogenesis of the diseases included in it. There are likely a variety of immunologic mechanisms among the distinct diseases in this category. (See 'Polymyositis' above.)

●Multiple immune components, including T cells, have been suggested as pathogenic factors in antisynthetase syndrome. (See 'Antisynthetase syndrome' above.)

●The mechanism of myofiber injury for all forms of IMNM is unknown, speculatively related to autoantibody binding to the surface of myofibers and subsequent complement-mediated cellular injury. Histologically, IMNM is characterized by scattered necrotic and regenerating fibers with relative absence of lymphocyte infiltration and expression of major histocompatability complex class 1 (MHC1) and membrane attack complex (MAC) on sarcolemma of non-necrotic fibers. (See 'Immune-mediated necrotizing myopathy' above.)

●Other forms of myositis have been described, some of which exhibit characteristic histologic features, including drug-induced myopathy, eosinophilic myositis, granulomatous myositis, and myositis associated with graft-versus-host disease. (See 'Other myopathies' above.)