Juvenile Myositis: Insights on Classification, Pathogenesis and Therapies over the Past 50 Years

Abstract

The juvenile idiopathic inflammatory myopathies (JIIM) are rare serious, multisystem autoimmune diseases affecting children and young people. Fifty years ago the prognosis for these conditions was poor, with high morbidity and mortality, and very few treatments were available. Since then, there has been major growth in our understanding of the pathogenesis and evidence-basis for treatment. Early descriptions considered juvenile dermatomyositis (JDM) to be essentially one disease. Our modern understanding has shown that that there are different types of JIIM, and that while JDM remains the most prevalent, even JDM itself is highly heterogenous. Much progress has been made possible through concomitant growth in the research community with a focus on JIIM, and the fact that the myositis research and clinical communities are highly collaborative internationally. Here, we review the growing understanding of the different types of childhood myositis, the developments in classification, driven in part by recognition of associations of particular phenotypes with specific myositis autoantibodies, the novel insights into pathogenesis of the JIIM, the therapeutic targets that these have revealed, and the evidence for, and development of, modern treatment care pathways, which have changed radically in recent decades. This has resulted in improved outcomes for children and young people with JIIM, and hope for further groundbreaking developments.

Insights into the underlying clinical features and patho-mechanisms of childhood- onset (juvenile) idiopathic inflammatory myopathies (JIIM) date back more than 50 years. Major progress in understanding the clinical features and pathogenesis of JIIM, and in the development of new therapies, has been made in recent decades, related to several key developments in research collaboration, registry development and research support. National registries for JIIM were established in the mid-1990s in the United States (U.S.) and United Kingdom (U.K.) and later in other countries, which contained significant patient numbers with detailed clinical data and linked biospecimens¹. These registries facilitated the advancement in our understanding of the clinical spectrum and epidemiology of JIIM, risk factors for poor outcomes, as well as pathogenesis and biomarkers, as detailed below. Collaborative research networks both across the whole life course in myositis as well as within the juvenile dermatomyositis (JDM) research community, were founded in the mid to late 1990s, including the International Myositis Assessment and Clinical Studies (IMACS) Group, the Childhood Arthritis and Rheumatology Research Alliance (CARRA), the Paediatric Rheumatology International Trials Organisation (PRINTO), and JDM working groups within both the American College of Rheumatology and the Paediatric Rheumatology European Society (PReS)². The international multispecialty collaborative research environment resulted in the development of validated core set measures and response criteria^3,4 which facilitated the first multi-center therapeutic trials⁵, standardization in the classification of myositis⁶, and the development of therapeutic guidelines and standardized treatment approaches^7–9. A biannual international multidisciplinary scientific meeting dedicated to all forms of myositis, the Global Conference on Myositis (GCOM), was established in 2015 and has grown to more than 700 attendees¹⁰.

Another key milestone has been the establishment of myositis-focused patient foundations such as CureJM and Myositis UK that have supported JIIM patients and raised funds for research. Founded in 2003 by parents and grandparents, the U.S.-based Cure JM Foundation started as a volunteer-driven non-profit dedicated to earlier diagnosis, finding better treatments and ultimately a cure for JM. They have funded over $30 million in research grants and programs, often funding young investigators, and supported the CARRA biorepository for JDM. Spurred by the parent leaders of Cure JM, Myositis and You was published in 2007 as comprehensive handbook for parents and trainees¹¹. Cure JM also established several centers for specialized care and research on JM, and more recently established a Clinical Care Network, connecting more than 60 children’s hospitals in the U.S. and Europe, in which pediatric rheumatologists are focused on delivery of excellent clinical care to JM patients. The charity Myositis UK (https://www.myositis.org.uk/) has been instrumental in supporting progress in research and towards better outcomes for people of all ages living with myositis. Myositis UK has not only provided vital funding for pilot work, (which led to far larger awards from government and other funders), but also supported multiple research events, family and patient days, and information sharing initiatives. Both foundations have supported travel bursaries to enable clinical science and research fellows to attend the international GCOM meeting, and Myositis UK has funded innovative speed funding grants for young investigators at GCOM.

Diagnosis and Classification.

Myositis in both children and adults was diagnosed for many years according to the Bohan and Peter classification criteria; they used a combination of clinical features (proximal weakness, characteristic DM rashes) and laboratory abnormalities comprising elevation in serum muscle enzymes, electromyography and inflammatory muscle histology¹². JDM was the only recognized childhood IIM; it was considered a separate disease from adult myositis, with more frequent calcinosis and vasculopathic features (like cutaneous and gastrointestinal ulceration), and a better prognosis (Figure 1A). In the early 1990s, several myositis autoantibodies were identified in adult patients and found to define phenotypes with distinct clinical features, responses to treatment, outcomes, and HLA risk alleles¹³ (Table). These same myositis autoantibodies were then reported in several children with myositis, with similar clinical features as in adults¹⁴, resulting in proposals to expand myositis classification to include both clinical and serologic subgroups and to be parallel in children and adults^{13, 15}. Re-classification was not favoured until additional myositis autoantibodies (anti-TIF1, -NXP2 and -MDA5) were identified and found to be present in a significant proportion of JDM patients. National registries in the U.S. and U.K. defined the array of clinicopathologic and myositis autoantibody subgroups, showing a large degree of overlap in phenotypic features with adult myositis – with some important distinctions (Figure 1B)^16–19.

Figure 1.

A. Classification of Myositis – THEN. This shows the classification of myositis in the early 1990s, with separation of clinical-pathologic and myositis autoantibody subgroups. Juvenile dermatomyositis was considered a distinct disease from the adult forms of myositis, other subgroups of juvenile myositis were not recognized, and the number of myositis autoantibodies identified was limited.

B. Classification of Juvenile Myositis – NOW. This represents the classification of myositis today, with integration of clinical subgroups and myositis autoantibodies to together define distinct phenotypes. Most of the subtypes present in adult myositis are now recognized in juvenile myositis, and classification is integrated across the age-span. Many more myositis autoantibodies and their associated phenotypes have been identified in recent years. Modified from^{4, 16, 119}.

Abbreviations: HMGCR, 3-Hydroxy-3-Methylglutaryl-CoA reductase; MDA5, Melanoma Differentiation-Associated Gene 5; NXP2, Nuclear Matrix Protein 2; RNP, Ribonucleoprotein; SAE, Small Ubiquitin-like Modifier Activating Enzyme; SRP, Signal Recognition Particle; TIF1, Transcription Intermediary Factor 1.

Table.

Key Milestones in Juvenile Myositis Research, Pathogenesis and Therapy

1891: Unverricht reports first recognized cases of adult DM

1952: ACTH therapy (corticosteroids) introduced as first effective medication for JDM

1966: Banker JDM autopsy series details natural history and primary pathologic/pathogenic features

1975: Bohan and Peter classification (diagnostic) criteria

1980: Methotrexate introduced as steroid sparing medication

1983: High dose corticosteroids are standard therapy

1993: IVIG efficacy established in adult DM

1994: Myositis autoantibodies and their phenotypes first recognized in children.

2002: Type I (αβ) and Type II (ᵧ) IFN expression discovered to be upregulated in JDM muscle

2010: First CARRA consensus treatment plan for JDM published towards standardization of initial therapy

2013: Rituximab trial published as first major clinical trial and first biologic therapy for JDM

2013-present: MYOGEN GWAS and ImmunoChip studies confirm HLA ancestral haplotype as major genetic risk factor, identify several other immune SNPs as risks for JDM

2016: PRINTO randomized trial demonstrates early introduction of Methotrexate superior to prednisone alone in new-onset JDM

2017: ACR-EULAR Myositis Response Criteria and Myositis Classification Criteria published; SHARE guidelines for JDM published and widely used across Europe

2018–2020: JAK inhibitors introduced in JDM

2021: FDA approval of IVIG for adult DM

2023 to present: Single-cell and omics studies identify novel pathogenic pathways (innate immunity, neutrophil, mitochondrial, PI3AKT and MAPK); Phase 3 trials in progress for biologic therapies for adult myositis; CAR-T therapy holds promise of remission

Abbreviations: ACR, American College of Rheumatology; ACTH, adrenocorticotropic hormone; CAR-T, chimeric antigen receptor T-cell therapy; CARRA, Childhood Arthritis and Rheumatology Research Alliance; DM, dermatomyositis; EULAR, European League Against Rheumatism; FDA, U.S. Food and Drug Administration; GWAS, Genome-Wide Association Study; HLA, Human Leukocyte Antigen; IFN, Interferon; IVIG, intravenous immunoglobulin; JAK, Janus kinase; JDM, juvenile dermatomyositis; JM, juvenile myositis; MAPK, mitogen-activated protein kinase; PI3K/AKT, phosphoinositide 3-kinase; PRINTO, Paediatric Rheumatology International Trials Organisation; SHARE, Single Hub and Access point for pediatric Rheumatology in Europe; SNP, single nucleotide polymorphism

The EULAR-ACR classification criteria were developed for myositis in 2017 as criteria to be used in research studies. JIIM was included, and the classification of JIIM and adult myositis is now combined into a single system⁶. These newer criteria include additional features in the patterns of muscle weakness, the presence of dysphagia and of anti-Jo1 autoantibodies, and further define muscle biopsy features. Of note, the EULAR-ACR IIM classification criteria are currently undergoing revision to potentially incorporate additional myositis autoantibodies, clinical-pathologic subgroups, and testing procedures, such as magnetic resonance imaging²⁰.

The advent of high throughput assays with improved reliability in detecting myositis autoantibodies has also led to routine clinical testing for myositis autoantibodies – not only to aid in diagnosis, but to understand associated clinical features and potential prognosis²¹. Currently, many different assays to test for myositis autoantibodies are available, and some carry a risk of false positive or false negative results; future work to standardize assays will be important²².

Myositis-specific autoantibodies (MSA) occur exclusively in patients with IIM, while myositis-associated autoantibodies (MAA) are also seen in overlap myositis and other systemic autoimmune diseases²³. Up to 70% of JIIM patients have at least one myositis autoantibody²⁴. The demographics, clinical and pathologic features, as well as therapeutic responses and outcomes for each autoantibody subset appear, to a degree, to be distinct and share similarities with the corresponding subset of adult IIM (Figure 1B). The role of MSAs in pathogenesis is also emerging, with demonstration of inhibition of their intracellular targets in vitro²⁵ and associations with distinct interferon (IFN) and other gene signature pathways²⁶; mouse models of some MSAs recapitulate pathologic and/or clinical features of disease²¹.

Anti-p155/140 (TIF-1ᵧ), the most common MSA in JIIM, has been detected in 18–30% of patients with JDM and JDM overlapping with another systemic autoimmune disease. Most children with this autoantibody have typical features of JDM, with widespread photosensitive rashes over the face, trunk, and extremities, as well as a higher frequency of cutaneous ulcerations, subcutaneous edema, and lipodystrophy (Figure 1B), and frequently a chronic illness course^{16, 17, 21}. Anti-TIF1ᵧ autoantibodies have also been associated with residential ultraviolet radiation exposure prior to diagnosis²⁷. Anti-TIF1ᵧ can also be associated with clinically-amyopathic JDM (CADM). In children, this autoantibody is not associated with malignancy, in contrast to adults. The reasons for this difference are unclear, but certain autoantibodies (including anti-Sp4, anti-CCAR1 and others) associated with anti-TIF1ᵧ appear to attenuate the occurrence of cancer in adults, while in JDM, these additional autoantibodies are also present and associated with milder muscle disease²⁸.

Anti-NXP2 autoantibodies are present in 15–25% of juvenile IIM patients, only in JDM and not in other subgroups. Anti-NXP2 is associated with severe weakness and muscle inflammation, muscle atrophy and cramping, joint contractures, mild skin rashes (only Gottron’s, heliotrope and malar rashes), and an increased frequency of calcinosis, although not in all studies (Figure 1B). In some cohorts internal organ involvement, including dysphagia and cardiac disease, are more frequent^29,30.

Anti-MDA5 autoantibodies are present in 6–8% of North American and European patients with JDM and 30% of Japanese children. This autoantibody is associated with mild muscle disease as well as frequent fevers, weight loss, adenopathy, arthritis, cutaneous ulcerations, and interstitial lung disease (ILD), which can be rapidly progressive (RP-LD) with associated high mortality^31,32 (Figure 1B). ILD and RP-LD are more frequent in Asian populations³³. Prognosis is good in patients without ILD³².

Anti-Mi2 autoantibodies, are present in up to 4% of JIIM patients, exclusively in JDM. They are seen more commonly in Hispanic patients in North America, and associated with more severe weakness and muscle inflammation, but also with a good response to corticosteroid and other immunosuppressive therapy^30,34.

Anti-synthetase autoantibodies are present in 2–5% of juvenile IIM, but up to 25% of adults with IIM. Anti-Jo1 (histidyl-tRNA synthetase) is the most common. These patients can have severe muscle weakness, with frequent polyarthritis, ILD, fever, Raynaud’s, and mechanic’s hands (Figure 1B). They often have a chronic illness course, with disease flares when corticosteroids are reduced, and increased mortality^30,35.

Two autoantibodies define immune-mediated necrotizing myopathy (IMNM), a form of JIIM that is associated with severe muscle weakness. Muscle biopsy may show myonecrosis, with myophagocytosis by infiltrating macrophages, along with features shared with other forms of JIIM including upregulation of MHC Class I and complement deposition on myofibers³⁶. Anti-signal recognition particle (SRP) autoantibodies are present in 2–4% of JIIM patients, most often African American individuals, and they generally have very severe proximal, axial and distal weakness, frequent falling episodes, muscle atrophy, markedly elevated serum creatine kinase (CK) levels (>10,000 IU/ml), increased cardiac involvement and restrictive lung disease, and may require a wheelchair or other devices due to severe functional disability³⁰ (Figure 1B). They generally do not have cutaneous features of JDM. They typically have a chronic illness course and require multiple immunosuppressive therapies. Patients with autoantibodies to 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR) also have severe proximal and distal weakness, falling episodes, highly elevated CK levels, frequent muscle atrophy, joint contractures, arthralgias, dysphagia, weight loss, and a chronic illness course^19,37. While anti-HMGCR autoantibodies in adults are frequently associated with use of statin medications or dietary exposure to statins, statin exposure has not been identified in the children.

MAAs are present in 11–36% of JIIM patients and have been associated with myositis overlapping with another systemic autoimmune disease in children and adult IIM patients (Figure 1B). In JIIM, MAAs overall are associated with increased frequency of Raynaud’s and ILD, a chronic illness course, and increased mortality, which increases with the number of MAAs present¹⁸. Among the MAAs, anti-Ro52 autoantibodies have been associated with worse ILD and more severe disease³⁸. Anti-Ro60 autoantibodies have been associated with panniculitis and weight loss. PM-Scl autoantibodies have been associated with higher CK levels and certain cutaneous manifestations, including sclerodactyly, mechanic’s hands, Raynaud’s, calcinosis and lipoatrophy. Other MAAs, including Ku and U3RNP, are sometimes seen in scleroderma-overlap myositis^17,18.

Areas of future research on myositis autoantibodies will aim to further define distinctive associated illness features among different geographically dispersed populations and races/ethnicities, to utilize novel technologies to identify myositis autoantibodies in patients who have no identified autoantibodies, and to further define the role of MSAs in pathogenesis. Reliable commercial assays for all autoantibody specificities are needed²². The use of sensitive ELISA and other quantitative methodologies to enable following titers over the illness course as biomarkers of disease activity and treatment response is another emerging area³⁹.

Pathogenesis.

An early description of muscle pathology by Banker and Victor⁴⁰ detailed tissue pathology in eight uniformly fatal cases of JDM. This study described key pathologic features, including early perivascular inflammation, vessel intima thickening, sometimes with occlusion, and even muscle infarcts. Early recognition that this necrotizing vasculopathy was present in muscle and other tissues was subsequently confirmed in many studies, found to correlate with prognosis⁴¹, and included in the Bohan and Peter classification criteria for JDM¹². More recently these features were incorporated in the ‘vascular domain’ of an internationally-validated consensus muscle biopsy score tool to assess JDM severity⁴². This cardinal pathological feature is thought to be associated with complement deposition on capillary endothelium⁴³; this pathology is consistent with more recent studies demonstrating vasculopathy in JDM, detected via endothelial biomarkers such as circulating endothelial cells (CEC) and endothelial-derived microparticles, as well as angiogenesis proteins, which are associated with IFN-driven pathology⁴⁴.

Early studies revealed increased expression of type I and II MHC (HLA) proteins on muscle fibers compared to age-matched healthy muscle tissue, detectable even early in disease in ~90% of JDM cases⁴⁵. Interestingly, over expression of MHC Class I protein specifically on muscle fibres led to myositis in a transgenic mouse model, with associated muscle inflammation and ER stress, and this model was more severe when the over-expression was initiated in young mice compared to adults^{46, 47}. Upregulation of Type I IFN gene expression was also prominently detected in muscle tissue in the first study of gene expression in JDM, in distinction to biopsies from Duchenne’s muscular dystrophy⁴⁸ (Table). Subsequent analysis of expression of the IFN-driven myxovirus-resistance protein A (MxA) in muscle biopsy tissue from 103 cases showed that the degree of upregulation of IFN-driven proteins correlates with muscle weakness and is observed across all MSA-specificities⁴⁹

IFNs drive a readily detectable transcriptional programme which includes multiple IFN-stimulated genes (ISG)⁵⁰, making this signature readily amenable to detection. Early demonstration of IFN-driven gene expression signatures in blood, muscle, and skin of both adult and juvenile^{51, 52} DM has been replicated in many studies using a variety of techniques, including measuring ISG scores in whole blood^{53, 54}, or sorted immune cells, analysing IFN-driven proteins in serum (such as Galectin-9, CXCL10)⁵⁵, or IFN proteins directly using the highly sensitive SIMOA^® assay⁵⁶. In the majority of these studies IFN-driven readouts correlate with muscle weakness, especially prior to treatment, as well as skin activity in patients with anti-TIF1ᵧ autoantibodies, and may predict flares^{51, 53, 54, 57}.

The IFN-driven signature is highly evident in muscle tissue both in bulk RNA analyses^{51, 58} and more recently spatial transcriptome studies^{59, 60}. The higher-powered cellular discrimination enabled by spatial transcriptome data has confirmed that muscle itself can be a source of IFN, in addition to infiltrating IFN-producing immune cells, which include T cells, NK cells and plasmacytoid dendritic cells (pDC)^{61, 62}. In additional to type I IFNs, growing evidence supports a role of type II interferon, IFNᵧ in JDM pathology, including direct detection⁶² and analysis of transcripts that are predominantly driven by IFNᵧ: CIITA and CXCL9⁶³.

Together these data, and early clinical efficacy data, have built a strong case for the use of IFN-blocking therapies, whether by blocking signalling through JAK/STATs, blockade of the IFN receptor, or targeting the cytokines themselves (see below). However, the immune system works as an orchestra with many parts interacting, not through one pathway only, and it is therefore not surprising that many immune processes have been implicated in the pathology of JDM. The dense early inflammatory infiltrate seen in JDM muscle includes T and B cells, monocyte/macrophages, and pDCs^{64, 65}. Peripheral blood memory B cells of JDM patients are expanded pre-treatment, and this expansion is driven in particular by CD24hiCD38hi immature B cells, which demonstrate abnormally low IL-10 production, that correlates with the IFN score and likely compromises their immunoregulatory function^66–68. Within the T cell compartment, a skew towards Th2 or CXCR5+Th17 cells has been observed, with reduced Th1 cells^{68, 69} within the CD4+ population. How altered function of blood immune cells reflect those in affected skin or muscle remains to be confirmed.

Transcriptional analysis of blood monocytes in JDM has revealed highly dysregulated mitochondrial biology, functionally altered oxidative phosphorylation and structural abnormalities of mitochondria⁷⁰. Ultra-structural studies demonstrate mitochondrial abnormalities in JDM muscle, which has been recently confirmed both in skin samples (obtained by ‘tape stripping’) by bulk RNAseq⁷¹, and in muscle biopsies by spatial transcriptomics⁵⁹. There is a complex bidirectional relationship between IFN and mitochondrial biology, with evidence that IFN alters mitochondrial function, but also that oxidized mtDNA released is a potent ligand for ‘danger’ recognition receptors and thereby can drive IFN production⁷². Recent evidence suggests that abnormal mitochondria may also be transferred between cells, a possible mechanism of perpetuating pathology⁷³. These studies open a new set of potential therapeutic targets which may prove valuable in ongoing disease activity not controlled by conventional treatment. In parallel with monocytes, neutrophils in JDM are also abnormal, associated with so called neutrophil extracellular traps (NETs)^{74, 75}, which also contain mitochondrial DNA. Serum proteome studies indicate neutrophil activation in JDM⁷⁶ and the blood biomarker MRP8/14 (also known as calprotectin, or S100A8/A9) is raised in serum in parallel with disease activity⁷⁷ as well as detectable within muscle, released by infiltrating myeloid cells. Additional pathways appear to be upregulated in chronically active JDM, including prominent dysregulation of p38/MAP kinase, phosphoinositide 3-kinase (PI3K)/AKT, and Il-1 signalling, as well as several cytokines and growth factors and their receptors⁷⁸.

The underlying genetic background which confers risk of myositis includes the human leukocyte antigen (HLA) or MHC region on chromosome 6. Large collaborative studies including both adult and childhood-onset IIM have facilitated many novel insights. Thus, the so-called ancestral haplotype, HLA A1-B8-DR3-DQ2 detected in early GWAS studies and subsequently confirmed is a risk factor for inflammatory myositis across all ages, in white populations⁷⁹ (Table). A detailed analysis of the leading amino acid position conferring risk identified subtle differences between adults and children, while MSA-positive subtype-specific genetics also vary by age of onset^{80, 81}. Several other risk loci for DM/JDM, including PTPN22, TYK2 (itself the target of the drug), IRF4 and C4 deficiency, are associated with myositis as well as other autoimmune diseases^82–84 Thus, many parts of the immune system likely play a role in the pathogenesis of juvenile IIM; the next challenge is to dissect which aspects are age-specific as well as define those pathways which correlate with specific MSAs or phenotypes, or predict an active course despite treatment.

Treatment and Prognosis.

The last 50 years have brought about tremendous changes in the treatment of JIIM, and a great deal of success, especially for children with JDM. So-called ancillary treatments (physical therapy, sun protection, proper dietary counselling, mental health support) are just as important now as they have been since the 1980s.

In a landmark advance, it was apparent in the early 1950s that corticosteroids were beneficial for patients with JDM, resulting in marked improvement in mortality⁸⁵ (Table). Early papers emphasized the benefits of high dose (up to 2.0 mg/kg /day in divided doses) and prolonged courses of daily oral corticosteroids, resulting in decrease in calcinosis and improved outcomes^{86, 87}. High doses of intravenous corticosteroids (e.g., 30 mg /kg methylprednisolone, maximum 1000 mg) at the outset of treatment likely leads to more rapid improvement, although there are many potential acute side effects⁸⁸. Longterm studies have not shown clear benefit of pulse methylprednisolone over time⁸⁹. Before combination therapy was widely practiced, it was universally observed that treatment with corticosteroids alone was associated with all the well-known undesirable side effects of long-term steroids (Figure 2A).

Figure 2.

A. Treatment Paradigm for Juvenile Dermatomyositis in 1980s- THEN.

Daily oral prednisone therapy was the mainstay of therapy. Second line therapies were limited, and only cyclophosphamide and plasmapheresis were available for severe patients.

B. Treatment Paradigm for Juvenile Myositis – NOW.

A standardized approach to initial therapy now includes daily oral prednisone in combination with methotrexate, with option to add intravenous methylprednisolone and/or intravenous gammaglobulin. A number of medications are available as second- and third-line therapies, including targeted biologics and drugs and use of combination therapies. Modified from¹²⁰.

Abbreviations: CAR-T, chimeric antigen receptor T-cell therapy; IFN-b, interferon-β; IV, intravenous; JAKi, jakinib; kg, killigram; m², square meters; mg, milligram; MoAb, monoclonal antibody

Accordingly, following demonstration of its efficacy as a second-line agent, methotrexate (MTX) was introduced as a steroid-sparing agent and proved effective when started at treatment outset⁹⁰. A subsequent randomized trial (RCT) showed the superior efficacy and tolerability of combination corticosteroids and methotrexate in newly diagnosed JDM patients⁹¹ (Table).

Intravenous immunoglobulin (IVIG) was first used in childhood myositis in the 1980s. IVIG has been adopted as a widely used, beneficial, therapy worldwide. Rigorous observational studies have demonstrated efficacy in children^92,93, and a recent RCT proved this in adults with DM⁹⁴ (Table).

Based on this experience, groups like the Childhood Arthritis and Rheumatology Research Alliance (CARRA)^8,9, and the Single Hub and Access point for pediatric Rheumatology in Europe (SHARE)⁷ have tried to standardize treatment regimens using available evidence and consensus (Figure 2B) (Table).

More recently, large case series have demonstrated the likely efficacy of mycophenolate mofetil (MMF)^{95, 96} in addition to, or as an alternative, to MTX. Combination drug therapy has also been more widely used to treat refractory disease and aide in sparing of corticosteroids (Figure 2B). Additionally, a sophisticated analysis has demonstrated efficacy of cyclophosphamide on JDM disease activity⁹⁷, which previously had traditionally been reserved primarily for ulcerative disease or ILD.

Targeted biologic therapies and small biologically-targeted drugs have been the focus of therapeutic development in the last 20 years⁹⁸ (Figure 2B). The evidence base for these agents is small, uncontrolled, and not necessarily convincing for all agents. Anti-tumor necrosis factor (TNF) monoclonal antibody therapies⁹⁹, have had some efficacy in case series of refractory patients. Rituximab, a B-lymphocyte depleting biologic therapy, has been studied in a large multicenter RCT of refractory JDM and DM/polymyositis (PM)⁵. While there was apparent efficacy in the sample of children studied, there was no statistical significance, which could be explained by the low power in the subgroup analysis¹⁰⁰ (Table). An open-label trial of abatacept suggested efficacy for refractory muscle and skin disease, with blinded measures, MRI and biomarkers including IFNs and chemokines, also showing improvement¹⁰¹.

Some of the newest therapies for childhood myositis are directed towards blocking the effects of IFN and other cytokines, primarily aimed at type I and II IFNs. Janus kinase inhibitors (jakinibs) block the effects of Type I and II IFNs by preventing the phosphorylation of STAT proteins (which, if phosphorylated, would drive ISG transcription). Worldwide there is a growing experience with jakinibs through case series and a small number of open-label clinical trials^{102, 103}, with a high proportion of refractory JDM patients showing good responses in skin and muscle disease in JDM and DM, as well as improvement in ILD and calcinosis in fewer cases. Jakiniibs have been beneficial for refractory JDM patients with anti-MDA5 autoantibodies, which are associated with high Type I IFN responses, in improving lung disease, as well as mortality in adult patients¹⁰². Most recently, monoclonal antibodies targeting IFNβ (dazukibart)^104,105, or the type I IFN receptor (anifrolumab)^106–108 have shown remarkable success in treatment of very refractory skin and muscle disease and in treatment of ulcerative disease in patients with JDM, and phase 3 trials are in progress in adult DM.

There is great excitement, now, about the potential for chimeric antigen receptor T-cell (CAR-T) therapies in the treatment of childhood myositis and juvenile lupus. To date the reported cases have used CAR-T directed against B cells (CD19)¹⁰⁹, or B cell maturation antigen-(BCMA)¹¹⁰ with marked success, including major clinical improvement and near remission in all cases (Table). The reported success seems greater than conventional biologic B cell targeted therapies like rituximab. It is thought that CD19-targeting CAR-T cells are effective against some tissue B cells and not just circulating B cells, and that the tissue effect may differentiate CAR-T cell potency.

While many of the therapies developed over the last 50 years are supported by research evidence from case-series or causal modelling, rather than RCTs, the outlook for a child with myositis in 2026 is very different than it was in 1966. While proper epidemiological studies had not been performed, early reports suggested that JIIM, particularly JDM, had a grim prognosis during the period before specific medication treatment was available. Reportedly one-third of patients died, one-third recovered with some sequelae, and one-third were chronically disabled¹¹¹. The course was often described as being monocyclic (in which disease completely remitted after 2 or 3 years of treatment), polycyclic (in which active disease recurrences occurred after periods of medication-free remission), or as chronic continuous⁸⁷.

More recently, several trajectories of disease have been described. In an inception cohort of JDM patients followed into adulthood, 3 latent trajectories were described¹¹². Ten percent of patients were very severe at onset, responded poorly to initial treatment, and remained quite active despite therapy. Over half had a moderate to severe presentation, but responded dramatically to initial treatment and had no, or very mild, continuing disease activity. The remainder had a mild to moderate presentation, but responded slowly to initial therapy, and continued to have ongoing mild disease activity despite therapy. In the latter group, it was often skin disease that persisted. The best predictor of continued disease activity was an incomplete response over the first 6 months of treatment. A large U.K. cohort identified similar disease trajectories in JDM, with time since diagnosis, presence of respiration abnormalities and lipodystrophy as predictors of a subgroup of 11% of patients with high disease activity at presentation and more persistent disease¹¹³. It has been shown by two groups that skin activity responds more slowly, and is more persistent, than muscle activity¹¹⁴.

Although most patients respond well to therapy, and mortality is now less than 5% (in most centers much lower), damage is common¹¹⁵. In a remarkable series of studies first reported in 2009¹¹⁶, a Norwegian group has examined an inception cohort of JDM patients into adulthood and demonstrated that children with JDM go on to have frequent organ damage, which is predicted by persistent disease activity at 6 months of treatment. They commonly have chronic weakness and may develop sub-clinical heart and lung damage, and reduced exercise tolerance in these long-term outcome studies. Approximately two-thirds have persistently active disease that is associated with persistent nailfold capillary abnormalities¹¹⁷. Over one-fourth have reduced bone mineral density and metabolic abnormalities, with hyperlipidemia and central adiposity.

Calcinosis is a potentially devastating long-term complication of the JDM sub-type of JIIM. It occurs in about 30% of patients in most series and is associated with a chronic disease course and longer disease duration¹¹⁵. In one inception cohort, the only baseline clinical predictor of calcinosis was more extensive vascular disease, as determined by underlying nailfold capillary vasculopathy¹¹⁸. There is no evidence-based treatment for calcinosis; however, aggressive treatment of the underlying disease often leads to regression of calcinosis.

The outcomes of JIIM have clearly improved remarkably in the last 50 years, but there are still many challenges faced by these patients. Further refinement of treatment approaches with the adaptation of consensus treat to target recommendations with the aim of rapid induction of inactive disease and resultant reduction of corticosteroids will hopefully accelerate further improved outcomes¹¹⁹.

Conclusions.

In conclusion, huge progress has been made in our understanding, treatment, development of prognostic tools, and knowledge of the long-term outcomes of juvenile myositis. This rapid growth of knowledge has been driven in part by collaborative networks and organizations, which have been synergistic across pediatric and adult research communities, between clinical and non-clinical researchers, and through strong international partnerships. Much remains to be done, to further understand pathogenesis, improve treatments, and test new therapeutic modalities and targets, but the future of the field is bright.