Updates on Juvenile Dermatomyositis from the Last Decade: Classification to Outcomes

INTRODUCTION

Juvenile dermatomyositis (JDM) is a rare heterogeneous immune-medicated condition with characteristic skin rashes, muscle weakness, vasculopathy, and other organ involvement. Myositis-specific autoantibodies (MSAs) help define clinical associations within subgroups of JDM.^1,2 The authors present key updates from the past 10 years on the classification, pathogenesis, MSA groups, assessment, and treatment of JDM.³

CLASSIFICATION

Historically, classification criteria for JDM have been limited to those of Bohan and Peter from 1975.⁴ These were not originally intended to be used, however, for JDM. Lundberg and colleagues⁵ recently conducted a large, international, consensus-based effort to develop new classification criteria for inflammatory myositis in both adults and children. The criteria assign scores to a variety of characteristics, such as presence of classic rash, age, or biopsy features, giving a total score, which can be used to determine the likelihood that a patient has an idiopathic inflammatory myopathy. Additional validation of the criteria in JDM suggest excellent sensitivity but specificity that is lower than previous criteria.⁶ The role of the new criteria for clinical research or diagnosis remains to be determined.

SEROLOGIC CLASSIFICATION OF JUVENILE DERMATOMYOSITIS

MSAs are found in 50% to 70% of juvenile myositis and have multiple clinical associations.^1,2 MSAs are present exclusively in patients with myositis, as summarized in Table 1. Myositis-associated autoantibodies (MAAs) are present in patients with myositis as well as in patients with other autoimmune diseases but have clinical associations in myositis.

Table 1.

Myositis-specific autoantibodies in juvenile myositis

MSA Group	Alias(es)	Frequency (%)	Cohort Sizes	Associations	References
Anti-TIF1	p155/140, TIF1-gamma	17.9–35.0	21–380	Associated with photosensitivity, Gottron papules, malar rash, V-, and shawl-sign rash, cuticular overgrowth, cutaneous ulceration, and chronic or polycyclic disease course	1,2,9,12,14
Anti-NXP2	MJ	15.1–23.8	21–380	Associated with muscle cramps, more weakness, dysphonia, calcinosis, hospitalization, younger age of onset, and less remision at 2 years	1,2,11,14
Anti-MDA5	CADM-140	6.1–23.8	13–453	Associated with oral and skin ulceration, amyopathic disease, ILD (sometimes RP-ILD), arthralgia/arthritis, fever, weight loss, adenopathy, and older age of diagnosis	2,7,10,12,14,15
Anti-synthetase	Jo-1, PL-7, PL-12, other aminoacyl-tRNA synthetases	0.0–5.1	21–380	Associated with ILD, arthralgia, less weakness, mechanic’s hands, lipoatrophy, and older age of onset	1,2,12–14
Anti-Mi-2	NuRD	0.0–3.9	21–380	Associated with malar rash, more muscle weakness including pharyngeal weakness, edema, less ILD, lower mortality, and more common in Hispanic patients	1,2,12–14
Anti-SRP		0.0–1.8	22–380	Associated with polymyositis (without rash), severe onset, more weakness including distal weakness, falling episodes, necrosis on muscle biopsy, Raynaud phenomenon, cardiac involvement, cutaneous ulceration, high CK, frequent hospitalization, wheelchair use, poor response to treatment, chronic disease course, and more common in African American patients.	1,2,12,13
Anti-HMGCR		1.1	380–440	Associated with severe proximal and distal weakness, necrosis on muscle biopsy, muscle atrophy, joint contractures, arthralgias, dysphagia, high CK, HLA DRB1*0701, poor response to medication, and chronic course. No previous statin medication exposure.	2,8
Anti-SAE		0.3–9.1	11–380	Too few cases with in juvenile myositis for any clear clinical associations	1,2,19

Abbreviations: CK, creatine kinase; NuRD, nucleosome remodeling deacetylase complex; RP-ILD, rapidly progressive ILD.

Myositis-specific Autoantibodies

In the past decade, several studies assessing clinical associations of MSAs have shown generally consistent findings. Anti-transcriptional intermediary factor 1 (TIF1) is the most common MSA, distinct from adult myositis in which anti-synthetase autoantibody is most common. The largest cohorts are based in North America (n = 365–454)^1,7–9 and the United Kingdom (n = 285–380),^2,10,11 with smaller cohorts from India^12,13 and Japan.^14,15 MSA-related clinical phenotype subgroups have been defined based on not only clinical features but also demographics and disease course (Table 1). Other distinctions are highlighted.

Environmental associations were also identified. Anti-p155/140 or TIF1 has an association with higher UV indices,¹⁶ consistent with the clinical association of anti-TIF1 with photosensitivity.^1,2,9,14 Higher UV index also was associated with lower odds of antinuclear matrix protein 2 (NXP2), also known as anti-MJ.¹⁶

In one study, the IFN score of anti-TIF1 autoantibody-positive JDM patients had higher correlation with skin activity than in JDM generally or other disease activity measures, even though anti-TIF1 patients did not have significantly higher skin disease activity.¹⁷ This suggests that biomarkers may have distinct associations by MSA group, which may point to differences in pathogenic mechanisms, but this needs further evaluation.

JDM patients with anti-Mi2 were less likely to remain on treatment despite more severe muscle biopsy findings, indicating they may be more responsive to treatment.¹⁸ Thus, although muscle biopsy general indicates disease severity, MSA group also should be considered when considering disease course and treatment.

For the anti–melanoma differentiation-associated gene 5 (MDA5) MSA group, there are notable differences in cohorts based on geography. In general, the anti-MDA5 MSA group has less weakness with more joint symptoms (arthritis and arthralgia), skin ulcerations, and systemic features (eg, fever). In Japanese JDM cohorts, however, anti-MDA5 antibody was more frequent (23.8%–38.5%)^14,15 and associated with rapidly progressive interstitial lung disease (ILD).¹⁵ In other cohorts, anti-MDA5 was less common (6.1%–7.7%), associated with ILD, but not associated with rapidly progressive ILD.^2,7,10,12

Necrotizing myopathy includes anti–signal recognition particle (SRP), which is associated with juvenile polymyositis or JPM (without skin findings) and the recently described anti–3-hydroxy-3-methylglutaryl-coenzyme A reductase (anti-HMGCR). Anti-HMGCR recently was detected in 1.1% of JDM, JPM, and necrotizing myositis of both North American and UK-based cohorts.^2,8 Distinct from adults, it has not been associated with previous statin exposure and has a different HLA-type association (DRB1*0701) compared with adult myositis.⁸

Anti–small ubiquitin-like modifier activating enzyme (anti-SAE) is another rare MSA in juvenile myositis. There is limited information on the few cases from 3 cohorts (1/374 North American cohort, 3/380 UK cohort, and 1/11 Japanese cohort), which is too limited to establish clear clinical correlations.^1,2,19

There can be differences in testing sensitivity and specificity by MSA testing method, so care should be taken when interpreting results.²⁰ This is an important area of continued study.

Myositis-associated Autoantibodies Updates

Anti-NT5C1A, a more recently identified MAA, was identified in adults with inclusion body myositis. Of 380 US-based juvenile myositis patients, 27% were found to have anti-NT5C1A with greater pulmonary symptoms at diagnosis, more frequent hospitalizations, and required more medication.²¹ Also, anti-Ro52 MAA was assessed in juvenile myositis (14% in North American cohort of 698) to be more frequent with anti-synthetase and anti-MDA5 MSA groups. In general, it was associated with ILD, chronic disease course, less remission, and more medications.²²

ASSESSMENT

Assessment of children with JDM for the purposes of clinical care or research is challenging due to the complexity of the illness. Consideration needs to be given to various aspects of muscle involvement, a wide range of skin manifestations, and involvement of multiple organ systems as well as other aspects of disease impact, including quality of life (QoL). Measurement tools in JDM are reviewed by Rider and colleagues.²³ New or updated/modified assessments discussed in this article are listed in Table 2. A key development in the understanding of the assessment of JDM was the advent of core sets, consensus-derived groups of outcome measures, which attempt to assess the full spectrum of disease manifestations. The core sets were developed independently by the International Myositis Assessment and Clinical Studies Group (IMACS)²⁴ and Pediatric Rheumatology International Trials Organization (PRINTO) groups²⁵ and are summarized in Table 3. Although this approach acknowledged the multidimensional nature of JDM, completion of all recommended core set measures is a time-consuming, challenging process that is difficult to use in a clinical context.

Table 2.

New and updated/modified juvenile dermatomyositis assessment tools discussed in this article

Measure	What is Assessed?	Author (Reference)
IMACS core set	Disease Activity	Miller et al,24 2001
PRINTO core set	Disease Activity	Ruperto et al,25 2003
Consensus core set	Comprehensive assessment	McCann et al,26 2014; McCann et al,27 2015; McCann et al,28 2014
Hybrid MMT-8/CMAS	Muscle strength/function	Varnier et al,29 2018
JDMAI	Muscle strength/function	Rosina et al,30 2019
Abbreviated MMT-8	Muscle strength	Rosina et al,31 2020
Whole-body MRI score	MRI muscle inflammation	Malattia et al,33 2014
Pelvic girdle MRI score	MRI muscle inflammation	Thyoka et al,34 2018
Muscle biopsy score	Muscle biopsy assessment	Wedderburn et al,36 2007; Varsani et al,37 2015
CAT	Skin disease activity	Huber et al,38 2008
CDASI	Skin disease activity	Klein et al,39 2008
DSSI	Skin disease activity	Klein et al,39 2008
DAS	Skin disease activity	Bode et al,40 2003
Inactive disease	Inactive disease	Lazarevic et al,47 2013; Almeida et al,48 2015
Definition of improvement	Response/im provement	Ruperto et al,49 2010
Response to therapy	Response/im provement	Rider et al,50–52 2017

Table 3.

Juvenile myositis disease activity core set measures

IMACS Core Set variables59	PRINTO Core Set variables60
Physician global disease activity (VAS or Likert)	Physician global disease activity (VAS or Likert)
Parent/Patient Global Disease Activity (VAS or Likert)	Parent/patient assessment of overall well-being (VAS or Likert)
CHAQ and CMAS	CHAQ
MMT	CMAS and MMT
Muscle enzymes (at least 2)	Muscle enzymes
Extraskeletal muscle disease activity (to be developed)	Global JDM disease activity tool

Consensus Core Data Set in Juvenile Dermatomyositis

To facilitate collaboration and standardization of data collection across international centers, McCann and colleagues²⁶ developed a consensus core data set. They developed a preliminary minimal data set using consensus methods. Subsequently, this data set was reviewed and refined through an international email-based Delphi survey, including both medical specialists and patients/parents, followed by a smaller consensus meeting of JDM experts using nominal group technique.²⁷ The resulting data set included 123 items and was renamed a “core” rather than “minimal” data set.²⁸ Several organizations, including the Childhood Arthritis and Rheumatology Research Alliance (CARRA), have adopted the core data set. Collecting a common data set in both clinical and research contexts will facilitate future collaboration between sites and research networks.

Changes to Core Set Tools

Recent work seeks to streamline the assessment process. Varnier and colleagues²⁹ published preliminary validation work of a hybrid tool, which combines the manual muscle test 8 (MMT 8) with 3 items from the Childhood Myositis Assessment Scale (CMAS) to give a score from 0 to 100. Their work supported face and content validity, construct validity, reliability, internal consistency, responsiveness and discriminant validity, while reducing completion time. Rosina and colleagues³⁰ reported on the development of a composite measure of disease activity in JDM, called the Juvenile DermatoMyositis Activity Index (JDMAI). The JDMAI incorporates physician and parent/patient global scales, an assessment of muscle strength and an assessment of skin disease activity. All 6 versions of the JDMAI demonstrated face and content validity, construct validity, internal consistency, responsiveness, and the ability to discriminate between disease states and parental satisfaction. Future validation work will allow selection of a final version. Rosina and colleagues’ group also has reported on the measurement properties of shortened versions of the MMT-8.³¹ Rather than testing 8 muscle groups, they studied versions using 6 and 4 muscle groups, documenting similar construct validity, internal consistency, discriminant validity and somewhat better responsiveness to the original tool. Together, these publications demonstrate ongoing attempts to improve assessment of children with JDM, maintaining validity while facilitating use in clinical settings.

Magnetic Resonance Imaging

The use of magnetic resonance imaging (MRI) to assess and document muscle involvement in children with JDM has become common, even though MRI is not included in any diagnostic criteria. For example, in the CARRA Legacy Registry, which included 483 children with JDM, 90% had MRI as part of their evaluation.³² Malattia and colleagues³³ reported on the use of whole-body MRI in 41 JDM patients and controls. They developed and validated a whole-body MRI score, demonstrating reliability, construct validity, and responsiveness, and showed the presence of muscle inflammation in muscles not clinically affected. More recently, Thyoka and colleagues refined a previous MRI scoring tool for children with JDM,³⁴ using pelvic girdle MRI,³⁵ demonstrating good inter-rater and intra-rater reliability. Unfortunately, MRI results from this study were not linked to weakness or disease activity, limiting further assessment of the clinical utility of the score.

Muscle Biopsy

The use of muscle biopsy in the assessment of JDM has declined, largely due to its invasiveness and replacement by other methods that document muscle involvement. Many experts, however, continue to advocate for its routine use. Wedderburn and colleagues³⁶ described a consensus-derived scoring tool for comprehensive assessment of muscle biopsies. A subsequent update and validation showed good intra-rater and inter-rater reliability and strong correlations with clinical measures of disease activity.³⁷ Its use allows standardization of muscle biopsy assessment in clinical and research contexts. Deakin and colleagues¹⁸ showed that muscle biopsy combined with myositis specific autoantibodies predicted clinically relevant outcomes, including future treatment status, suggesting that muscle biopsy could be part of an assessment to tailor therapy for individual patients.

Assessment of Skin Disease in Juvenile Dermatomyositis

The assessment of skin disease in JDM is complicated by the use of multiple tools, including the Cutaneous Assessment Tool (CAT),³⁸ the Cutaneous Dermatomyositis Disease Activity and Severity Index (CDASI),³⁹ the Dermatomyositis Skin Severity Index (DSSI),³⁹ and the Disease Activity Score (DAS),⁴⁰ with little consensus about which tool is optimal. Given the number of patients with persistent skin disease long after resolution of muscle disease, the lack of consensus on assessment negatively affects the study of effective therapies. Several recent studies have attempted to address this lack of consensus.

Tiao and colleagues⁴¹ compared the CDASI and the CAT when used by pediatric dermatologists, rheumatologists, and neurologists in children with JDM. They found that both tools had good inter-rater reliability and correlations with other measures of activity and damage when used by dermatologists and rheumatologists, although performance was worse when used by neurologists. Both tools also were compared in adults with dermatomyositis (DM), with demonstration of good inter-rater and intra-rater reliability, construct validity, and responsiveness.⁴² In the adult study, including 10 dermatologists, the CDASI was preferred. A review of the CDASI, primarily in adults with DM, recently has been published.⁴³ Determination of the optimal tool to assess skin disease in children remains an area in need of consensus.

Quality of Life and Other Measures

There is increased recognition of the impact of JDM beyond physical manifestations like skin and muscle disease. Apaz and colleagues⁴⁴ reported on QoL, as assessed by the Child Health Questionnaire, in an international cohort of children with JDM. They found that QoL was impaired in children with JDM, particularly in the physical domain, compared with healthy controls, with baseline disability and disease duration predicting lower scores 6 months later. Tory and colleagues⁴⁵ surveyed health care providers and families of children with JDM to identify the most important components of high-quality care. They found that although patients and families generally gave higher scores than physicians, there was consensus on the 5 most important items: QoL, timely diagnosis, access to rheumatology care, normalization of function/strength, and ability for self-care. In another publication, Tory and colleagues⁴⁶ studied discordance between physician and patient/family global visual analog scales (VASs). They found that nearly 40% of the global VAS scores in a large registry were discordant, defined as a difference of greater than or equal to 2 points on a 10-point scale. More work is needed to understand the reasons for discordance between physician and patient/family global assessments and the impact on patient care.

Inactive Disease

Achieving inactive disease, either on or off medications, is an important goal in the management of JDM. Lazarevic and colleagues⁴⁷ used a large, international cohort of 275 children with JDM to develop and validate a data-driven definition (meeting at least 3 of 4 criteria: creatine phosphokinase (CPK) less than or equal to 150, CMAS greater than or equal to 48, MMT8 greater than or equal to 78, and physician global VAS less than or equal to 0.2). The definition, however, may overemphasize the importance of muscle disease, because a patient with marked persistent skin involvement or other active extramuscular disease would meet the definition of inactive disease if they had normal strength and normal muscle enzymes.⁴⁸ Almeida and colleagues⁴⁸ proposed an amendment whereby the physician global VAS is an essential criterion. Further work is needed to develop a more acceptable definition.

Definition of Improvement/Clinical Response in Juvenile Dermatomyositis

Defining improvement or response to therapy is critical to conducting clinical trials. Ruperto and colleagues⁴⁹ initially reported provisional response criteria for JDM. Using previously established core set variables for JDM, the criteria defined permitted percentages of improvement and worsening for a patient to be “improved.” More recently, an international effort developed new criteria for response to therapy in JDM.^50–52 The criteria generate a continuous total improvement score, based on points assigned for the absolute percent change of each of the core set variables. This approach allows for the definition of minimal, moderate, and major improvement, permitting quantification of the degree of response. The response calculator can be found online at https://www.niehs.nih.gov/research/resources/imacs/response_criteria/index.cfm.

PATHOGENESIS

Genetics

Although the etiology of JDM is not fully understood, genetic and environmental risk factors are key contributors. One study found that approximately half of JDM families had at least 1 other family member with autoimmune disease, most commonly type 1 diabetes mellitus or systemic lupus erythematosus.⁵³ Genome-wide association studies of single-nucleotide polymorphisms (SNPs) comparing adult DM and JDM to controls found multiple major histocompatibility complex (MHC) region signals, strongest at HLA 8.1 ancestral haplotype, with HLA DRB1*03:01 strongest in adult and JDM. Three SNPs also were more common from PLCL1, BLK, and CCL21, as has been seen in other autoimmune diseases.^54,55 Lower C4a gene copy number (deficiency) is more common in JDM than controls, particularly among those with an HLA DR3-positive background.⁵⁶ Genetic changes have been validated as an important component in JDM, identifying variants associated with immune response, particularly MHC class II, B cell signaling, and C4.

Environment

Because photosensitivity is a common manifestation of disease and flares often occur after sun exposure, UV radiation is considered an environmental trigger of JDM.⁵⁷ Higher UV index was associated more strongly with JDM and more prominently among those positive for anti-TIF1 MSA compared with juvenile myositis without skin involvement (JPM).¹⁶ Higher UV index in the month prior to symptom onset was associated with increased odds of developing calcinosis in non–African Americans with JDM.⁵⁸ When matched to healthy control mothers, maternal exposures to school chalk dust, gasoline vapor, smoking, and carbon monoxide during the third trimester were significant risk factors for having a child with JDM.⁵⁹ Direct UV exposure and maternal toxic exposures (chalk dust, gasoline vapor, smoking, and carbon monoxide) are environmental triggers associated with JDM.

Muscle Pathology

Perifascicular atrophy (PFA) is a characteristic finding on muscle biopsy in DM, with or without clinical myopathy.⁶⁰ In early DM without PFA, there already is focal capillary depletion found on biopsy.⁶¹ In chronic DM, biopsies show a combination of chronic ischemia with replacement of normal capillaries by abnormal lumens and thickened walls as well as neoangiogenesis.⁶⁰ RIG-I, an IFN-response gene part of innate immune system response, is overexpressed in areas of PFA. With hypoxia, RIG-I expression and type I IFN increase, indicating a connection between hypoxia, IFN, and PFA.⁶² When compared with adult myositis biopsies, JDM biopsies show predominantly hypoxia-driven pathology in perifascicular areas with atrophy, regional loss of capillaries, and less IFN-inducible gene expression.⁶³ Decreased vascular density, ischemia, cell death, neoangiogenesis, and IFN expression all contribute to PFA, a characteristic finding of JDM muscle biopsies.

Interferon

An increased interferon (IFN) regulated signature is present in JDM.⁶⁴ Human myogenic precursor cells from JDM patients’ muscle biopsies show an angiogenic gene expression signature recapitulated with type I IFN treatment, indicating JDM’s proangiogenic features may be type I IFN driven.⁶⁵ Galectin-9 and CXCL10, IFN-stimulated proteins, were validated as reliable biomarkers for disease activity in multiple JDM cohorts, showing elevated or rising levels prior to flare in longitudinal analysis.⁶⁶ Weakness and joint disease activity were the best predictors of a high IFN score from blood in JDM.¹⁷ Untreated JDM muscle biopsies had higher type I and type II IFN scores, both associated with muscle biopsy pathology and higher disease activity, whereas only type II IFN scores correlated with longer time to clinically inactive disease.⁶⁷ A UK group found MxA (myxovirus-resistance protein A) staining of JDM muscle biopsies by immunohistochemistry correlated inversely with muscle strength measurements.⁶⁸ Type I and type II IFN signatures are high in untreated JDM blood and muscle, correlating with disease activity and angiogenic changes.

Endothelial Markers

Endothelial dysfunction and vasculopathy are prominent in JDM. Low soluble intercellular adhesion molecule 1 (ICAM) and high endoglin levels are associated with increased vasculopathy assessed by nailfold capillary end-row loop scoring.⁶⁹ Endothelial injury markers (von Willebrand factor (vWF) antigen and circulating endothelial cells) correlate with extramuscular activity, whereas circulating endothelial progenitor cells negatively correlate with muscle activity.⁷⁰ Endothelial activation with elevated vascular cell adhesion molecule (VCAM)-1 was higher with shorter duration of untreated JDM (early in disease) in JDM muscle biopsy.⁷¹ Peripherally, endothelial activation (VCAM-1) and inhibition of angiogenesis (angiopoietin-2, soluble vascular endothelial growth factor receptor 1 (VEGFR-1)) also relate to JDM disease activity.⁷² As described previously, vasculopathy is a prominent feature of JDM with multiple endothelial markers that correlate to disease activity.

Immune Cells

Immature transitional B cells are expanded in pretreatment baseline JDM peripheral blood correlating with disease activity.⁷³ Genes related to B-cell survival, including BAFF, correlate with disease activity.⁷⁴ Increased plasmablasts correlate with some T-helper subsets and disease activity in JDM.⁷⁵ Although peripheral T-regulatory cells were found in the same proportion as controls, those cells from active JDM patients were found to be functionally impaired.⁷⁶ There are multiple immune cell subset changes in JDM, including B-cell and T-cell differences in level or function, which correlate with disease activity.

TREATMENT

There are few clinical trials to inform optimal treatment of JDM; therefore, treatment is largely based on empiric and expert opinion. Consensus recommendations and key clinical trial results are summarized.

Consensus Recommendations

Over the past decade, CARRA,^77–79 Single Hub and Access point for paediatric Rheumatology in Europe (SHARE),⁸⁰ the JDM working group of the Society for Pediatric Rheumatology⁸¹ and others have published standardized treatment approaches. Initial therapy for the average patient presenting with JDM typically includes a combination of steroids and methotrexate, with escalation or modification of therapy for resistant disease. See Figs. 1 and 2 for summaries of the CARRA and SHARE treatment recommendations.

Fig. 1.

Recommended consensus treatment plans from CARRA, for JDM patients with moderate, skin-predominant, and skin-resistant disease. IV: intravenous, IG: immunoglobulin. a: Lesser of 15 mg/m² or 1 mg/kg (max 40 mg), b: max 60mg, c: optional, d: max 70 grams, e: max 400mg, f: max 1500mg bid, g: higher doses based on toxicity, efficacy

Fig. 2.

Recommendations from SHARE, for treatment of JDM, ranging from mild to severe disease. * such as major organ involvement / extensive ulcerative skin disease; ** improvement based on clinical opinion, DMARD: disease modifying antirheumatic drug, MMF: mycophenolate mofetil, MTX: methotrexate

Clinical Trials

The Rituximab in Myositis trial was a randomized, double blind, placebo-phase trial⁸² for treatment of refractory myositis, which included 48 children. Although the study did not meet the primary outcome and definition of improvement, post hoc analyses suggest predictors of rituximab response included anti–Jo-1 and anti–Mi-2 or other anti-bodies compared with autoantibody negative patients.⁸³ The findings suggest a role for B-cell–depleting therapy for subsets of patients with JDM.

The PRINTO Network completed an international, multicenter, randomized, open-label superiority trial using steroids with or without methotrexate or cyclosporine,⁸⁴ enrolling 139 newly diagnosed JDM patients from 54 centers and 22 countries. The study confirmed the combination treatment with steroids and either methotrexate or cyclosporine to be superior to steroid monotherapy and methotrexate was more effective with less adverse events compared with cyclosporine.

Other Immunomodulatory Therapies

Intravenous immunoglobulin (IVIG) commonly is used as adjunctive treatment of JDM, but its efficacy is not well studied. A retrospective inception cohort of 78 Canadian JDM patients was assessed⁸⁵ to determine whether IVIG-treated patients had less disease activity and achieved inactive disease sooner than patients not treated with IVIG. Marginal structural modeling showed IVIG-treated patients started with greater disease activity compared with patients not treated with IVIG. After controlling for confounding, IVIG-treated patients maintained similar or lower disease activity compared with controls.

Mycophenolate mofetil is used to treat JDM patients with resistant and refractory disease. A small case series of 8 patients published in 2012⁸⁶ suggested mycophenolate mofetil was safe and efficacious with improvement in strength and the ability to wean steroids.

Cyclophosphamide typically is reserved for severe or resistant disease, with limited published efficacy data. Cyclophosphamide was used to treat 19 steroid-resistant JDM patients from a single Japanese hospital⁸⁷ and considered effective, including 4 patients whose interstitial pneumonitis resolved with treatment. However, 8 patients with calcinosis progressed without improvement. 56 patients from a UK JDM cohort received cyclophosphamide and showed improvement in global, skin, and muscle disease activity, with the greatest benefit in skin disease and global disease activity at 12 months.⁸⁸

The use of biologic agents to treat refractory JDM is evolving. Two small studies suggested that etanercept was not beneficial and could cause worsening of disease.^89,90 Other anti-TNF therapies, however, show promise for resistant disease. A retrospective analysis of 60 JDM patients treated with adalimumab or infliximab in the United Kingdom demonstrated benefit,⁹¹ including improvement of calcinosis. Smaller case reports and series also suggest anti-TNF therapies may be useful in resistant cases.^92,93

Single case reports of the benefit of abatacept in the treatment of JDM with associated calcinosis have been reported^44,94 and a larger pilot trial is under way (NCT02594735).

The use of Janus kinase inhibitors (JAKi) to treat JDM is increasing.^95–98 In a 2020 case series, of 25 refractory JDM patients treated with JAKi, 96% (24/25) reported improvement and 66.7% (16/24) had complete resolution of rash.^99,100 Kim and colleagues⁹⁵ reported 4 patients with chronically active JDM in a compassionate use study of baricitinib, which also supported JAKi efficacy, according to physician, patient, and extramuscular VAS, as well as skin scoring measures, biomarker response, and MRI.

Exercise

JDM patients may have long-term issues with exercise intolerance, muscle weakness, and fatigue, despite aggressive treatment and apparent clinical remission.^101,102 A growing body of evidence suggests a role for exercise training as an important non-pharmacologic adjunct treatment of JDM,^103–107 without causing flares in disease.

LONG-TERM OUTCOMES

Long-term outcomes reported from several international JDM cohorts^108–116 over the past decade continue to show suboptimal outcomes and long-term morbidity. These patient cohorts, followed for 3.7 to nearly 17 years, have chronic or polycyclic disease course in up to two-thirds of patients, suggesting that monocyclic disease course and remission are uncommon.

Although 60% of PRINTO’s 490 patient cohort had no impairment (Childhood Health Assessment Questionnaire [CHAQ] score = 0) after a mean disease duration of 7.7 years (range 2–25.2 y), approximately 7% had severe functional impairment (CHAQ score >1.5).¹⁰⁸ Adult JDM patients from 2 US adult referral subspecialty centers (n = 49)¹⁰⁹ had mild to moderate disability (median American College of Rheumatology [ACR] functional class 2, Health Assessment Questionnaire 0.4). More than half of JDM patients have reduced endurance and ongoing weakness,^108,113 with up to 10% reporting severe impairment¹¹³ at long-term follow-up, after a median of at least 6 years disease duration (range 2–25.2 y)^108,113 in 2 cohorts. These findings suggest that although some JDM patients have good functional outcomes, a subset of patients continue to have significant impairment due to JDM.

Cumulative damage at long-term follow-up, after a median of at least 6 years (range 2–25.2 years) disease duration,^108,113 is common, reported in up to 69% of patients.^108,113 Cutaneous damage also is frequent in these published JDM cohorts, including calcinosis and lipodystrophy, reported in up to 2% to 60% and 10% to 33% of patient, respectively.^{108,109,113,114} Long-term musculoskeletal damage based on MRI findings is present in up to 52% of patients¹¹⁰ and joint contractures in 18% to 63%.^108,109 There also is long-term growth delay in 8% to 40% of long-term cohorts followed for a median of at least 6 years (range 2–25 years).^{108,109,114,115} In summary, the long-term damage described in JDM patients likely is related to many factors, including disease activity as well as medication toxicity from corticosteroids. The high frequency of damage in these cohorts point to the urgent need for better JDM treatments.

Although mortality rates for JDM have improved since the 1960s, when it was reported in one-third of patients,¹¹⁷ current rates remain unacceptably high. The JDM specific mortality rate from the National Institutes of Health natural history questionnaire–based protocols administered between 1989 and 2011 was estimated at 2.4%, most commonly attributed to ILD.¹¹⁸ In other cohorts, death was attributed to severe JDM manifestations, including cutaneous ulceration with infections, gastrointestinal vasculitis, and lung disease, including ILD and respiratory failure.^{114–116,118} The dramatic differences in mortality ranging from 2¹¹⁸ to 11%,^114–116 suggest not only differences in patient populations, but disparities in access to care and treatment.

RISK FACTORS FOR WORSE OUTCOMES

The trajectory of disease is variable in JDM and it is challenging to predict long-term patient outcomes. Predicting treatment resistance or poor outcomes earlier in the course of disease could help guide treatment decisions to improve long-term prognosis.

There are inconsistent reports of age-based risk for worse outcomes in several cohorts that suggest possible regional or environmental differences contributing to worse outcome, and additional study is necessary.^{109,112,118,119}

Longer disease duration, including chronic or continuous disease course, predicted long-term damage and worse outcomes, including functional impairment and death, in several cohorts.^{108,109,112,113,118} Similarly, calcinosis is reported to be less common in patients with a monocyclic course.¹¹⁵ High disease activity, according to muscle and total DAS scores, 1 year after diagnosis predicted poor outcomes in a Norwegian cohort.¹¹⁰ Worse baseline ACR functional class and baseline skin disease and severity (including erythroderma, heliotrope rash, and shawl sign) also were found to predict damage and risk for mortality in JDM.^109,118 These findings suggest that patients with delayed or severe disease at diagnosis or persistent disease activity at 1 year may benefit from proactive, aggressive treatment to achieve rapid disease control and remission that may help limit unsatisfactory outcomes.

Ravelli and colleagues¹⁰⁸ reported female patients had more long-term functional impairment and worse health-related QoL. The JDM CARRA Registry found that higher mean UV radiation exposure 1 month prior to disease onset increased calcinosis risk and varied by race.⁵⁸ Decreased end row loops (ERLs) in nail fold capillaries at JDM diagnosis was associated with longer untreated disease and more active skin disease compared with children with normal ERLs.¹²⁰ The UK Juvenile Dermatomyositis Cohort and Biomarker Study found differences in muscle biopsy scores when comparing MSA subgroups and increased severity of the histopathologic features predictive of longer treatment duration.¹⁸ Weight loss was found an early illness feature associated with mortality.¹¹⁸ These findings suggest that the clinical history, physical examination at onset, and diagnostic work-up may be useful to inform treatment recommendations and help to improve future long-term outcomes.

Earlier prognostication with modifications in therapy tailored to each patients’ risk for worse outcome may be possible if the current wide-ranging predictors are validated in future studies. Additional research is needed to identify more effective treatment approaches targeted toward these patients at risk for worse outcomes.

SUMMARY

JDM is a heterogeneous disease with new classification criteria and updates in MSA and MAA groups, helping define clinical subgroups. There are many validated assessment tools for assessing disease activity in JDM. Future studies will optimize the tools, improving feasibility of use in clinical and research contexts. Genetic and environmental risk factors, mechanisms of muscle pathology, role of IFN, vascular markers, and changes in immune cells provide insights to JDM pathogenesis. Outcomes have improved, but chronic disease activity, damage, and mortality highlight the need for improved predictors of outcome and treatment efficacy. Increased international collaboration of patients, clinicians, scientists, and other key stakeholders in patient registries and biorepositories as well as clinical trials may overcome current barriers that exist for improving research, treatment, and outcomes for this rare condition.

KEY POINTS.

• New classification criteria for juvenile dermatomyositis (JDM) have been developed, but their role remains unclear. New serologic classifications that define distinct subgroups within juvenile myositis may be more useful in practice.

• There are many validated assessment tools available to assess disease activity in JDM. Future studies will optimize these tools and improve feasibility in clinical and research contexts.

• Genetic and environmental risk factors, causes of perifascicular atrophy in muscle, and vascular dysfunction provide new insights into the pathogenesis of JDM. Biomarkers related to endothelial function, interferon, and immune cells also have shown correlation to disease activity.

• Outcomes of children with JDM have improved with immunomodulatory therapies but many patients have persistent disease sequelae.

• An expanding repertoire of medications is helpful in subsets of JDM patients, but additional investigation to identify optimal individualized treatment regimens is essential to improve future outcomes in JDM

Clinical Care Points.

• Validated tools should be used for the clinical assessment of children with JDM

• Identification of specific JDM phenotypes using MSA and clinical features may help to predict course and outcome and influence treatment decisions

• Prompt diagnosis and treatment are needed to improve long-term outcomes in JDM

• Guidelines and recommendations developed through expert opinion and consensus should be used to treat JDM patients