Clinical features and disease course of patients with juvenile dermatomyositis


  • Peter J. Gowdie,

    1. Department of General Medicine, Rheumatology Service, Royal Children's Hospital, Melbourne, Victoria, Australia
    Search for more papers by this author
  • Roger C. Allen,

    1. Department of General Medicine, Rheumatology Service, Royal Children's Hospital, Melbourne, Victoria, Australia
    Search for more papers by this author
  • Andrew J. Kornberg,

    1. Department of Neurology, Royal Children's Hospital, Melbourne, Victoria, Australia
    Search for more papers by this author
  • Jonathan D. Akikusa

    Corresponding author
    1. Department of General Medicine, Rheumatology Service, Royal Children's Hospital, Melbourne, Victoria, Australia
    2. Murdoch Children's Research Institute, Royal Children's Hospital, Melbourne, Victoria, Australia
    • Correspondence: Dr Jonathan D. Akikusa, The Royal Children's Hospital, Flemington Road, Parkville, Victoria 3152, Australia.


    Search for more papers by this author



To describe the clinical features and course of a cohort of patients with juvenile dermatomyositis (JDM) at a tertiary referral pediatric centre in Australia and examine changes in diagnostic and therapeutic approach over time.


Retrospective review of patients diagnosed with JDM at the Royal Children's Hospital, Melbourne, between 1989 and 2010.


Fifty-seven patients were identified. The female : male ratio was 2 : 1 and median age at diagnosis was 7.1 years (2.2–15.3). At diagnosis, 95% had weakness, all had typical rash and 68% had nailfold capillary changes. Calcinosis was not present in any patients at diagnosis and occurred in 18% over time. Creatine kinase, lactate dehydrogenase, aspartate aminotransferase, alanine aminotransferase and aldolase levels were abnormal in 65%, 92%, 88%, 58% and 100%, respectively. Magnetic resonance imaging (MRI) was abnormal in 97% of patients, electomyograph (EMG) in 83% and muscle biopsy in all four patients in whom it was performed. MRI was used in 86% (24/28) of patients diagnosed after 2000. Muscle biopsy was used in four and EMG in no patients over the same period. Treatment used throughout the disease course included oral steroids (93%), high-dose pulse intravenous steroids (82%), methotrexate (63%), intravenous immunoglobulin (32%) and cyclosporin (18%). The disease was monophasic in 46.7% (21/45), polyphasic in 17.7% (8/45) and chronic in 35.5% (16/45).


Australian patients with JDM have similar characteristics to previously described cohorts. In practice, MRI has replaced the invasive diagnostic tests included in the Bohan and Peter criteria for the diagnosis of JDM. The early use of disease-modifying anti-rheumatic drugs has become the most common treatment approach.


Juvenile dermatomyositis (JDM) is a rare systemic vasculopathy of unknown aetiology. It is the commonest of the idiopathic inflammatory myopathies of childhood, comprising 85% of cases.[1, 2] It has an annual incidence estimated to range between 1.9 and 4.1 per million children.[3, 4]

Clinically, JDM is characterized by muscle weakness and typical skin involvement. It may also involve multiple other systems, including the gastrointestinal tract, heart, lungs, kidneys and eyes. The diagnosis of JDM is based on criteria first proposed by Bohan and Peter in 1975.[5, 6] These criteria are: proximal muscle weakness, characteristic rash, raised muscle enzymes and typical electromyography (EMG) and muscle biopsy changes. In recent years magnetic resonance imaging (MRI) has played an increasingly important role in the diagnosis of inflammatory muscle disease and in many situations has obviated the need for invasive procedures such as EMG and muscle biopsy.[7]

Previous studies have described the clinical features and course of large JDM cohorts in North America, Europe, South America, Saudi Arabia and Japan. To our knowledge, there is only one other Australasian study that describes a cohort of patients with JDM.[8] The aim of this study was to describe the clinical features, complications, course and treatment of JDM at an Australian tertiary referral centre over the past two decades.

Patients and Methods

A retrospective chart review was conducted of all patients diagnosed with JDM at the Royal Children's Hospital (RCH) in Melbourne between 1989 and 2010. The study was approved by the RCH Human Research Ethics Committee. Patients were identified by two search strategies. The first involved a search of the hospital medical records database to identify patients discharged from the hospital between January 1989 and June 2010 with an International Classification of Diseases 9th or 10th edition (ICD-9 or ICD-10) code potentially compatible with the diagnosis of JDM. The ICD-9 codes used were 710.3 (Dermatomyositis) and 710.4 (Polymyositis) and the ICD 10 codes used were M33.0 (Juvenile Dermatomyositis), M33.1 (Other Dermatomyositis), M33.2 (Polymyositis) and M33.9 (Dermatopolymyositis, unspecified). The second search method involved interrogation of the Rheumatology Department's independent electronic database to search for patients assigned a diagnosis of JDM over the same period.

The charts of all patients identified were reviewed by a single reviewer (PG) and information concerning patient demographics, treating team, clinical features at onset and throughout the course of the illness, investigation results, and therapy were entered into an electronic database.

Patients were included in the study if they met the Bohan and Peter[6] criteria for definite, probable or possible JDM. Additionally, to be included patients had to have been managed at RCH throughout the course of their illness and have had at least 3 months of follow-up.

For the purposes of describing disease course, disease remission was defined as the absence of both rash and active myositis (i.e. normal muscle enzymes and normal muscle strength) maintained for a minimum of 6 months off immunosuppressive therapy. Normal muscle strength was defined as per the examination by the primary physician involved in the patient's care or as demonstrated on the Childhood Myositis Assessment Scale (CMAS) performed by a physiotherapist. The date of remission was calculated as the first date the patient was off all immunosuppressive therapy. Disease course was divided into three groups according to patterns of active and inactive disease: monophasic, polyphasic and chronic, based on previous descriptions in the literature.[7-9] A monophasic course was defined as remission of disease within 36 months of diagnosis without relapse thereafter. Polyphasic course was defined as remission followed by relapse of disease at any time point and a chronic course was persistent evidence of disease 36 months after diagnosis. When follow-up of patients was less than 36 months, the course of disease was unspecified. Relapse was defined as new evidence of disease activity (active myositis or rash) following at least 6 months of remission. Clinical features at onset were defined as those symptoms and signs documented at the time of diagnosis. Treatment at onset was defined as treatment commenced within 4 weeks of diagnosis. Second-line therapy was defined as any immunomodulatory agent used other than steroids.


Fifty-seven patients were identified, 38 (67%) were female. The median age at diagnosis was 7.1 years (range: 2.2–15.3; Fig. 1). The median duration of symptoms prior to diagnosis was 2.8 months (range: 0.7–20.5). The median length of follow-up was 4.0 years and the median age at last clinic visit was 13.2 years.

Figure 1.

Age and sex distribution of the juvenile dermatomyositis cohort.

Diagnosis of JDM

Of the 57 patients, 40% had ‘definite JDM’ (23/57), 56% had ‘probable JDM’ (32/57) and two patients (4%) had ‘possible JDM’ according to Bohan and Peter criteria. Eighty-eight percent of ‘probable JDM’ patients (28/32) had one or more of: abnormal MRI; nailfold capillary changes; calcinosis; or dysphonia/dysphagia. Of the two patients with ‘possible JDM’, one had typical JDM rash, abnormal nailfold capillaroscopy and muscle enzyme abnormalities, but normal muscle strength. Muscle biopsy and EMG were not performed; however, MRI demonstrated typical features of myositis. The second had characteristic JDM rash and weakness but normal creatine kinase (CK) and muscle biopsy. EMG was not performed; however, MRI was consistent with myositis.

Clinical features at diagnosis and course of illness

The clinical features of the 57 patients at diagnosis and at any time during follow-up are presented in Table 1. Ninety-five percent presented with clinically discernible weakness. Of the three patients without apparent weakness at onset of disease, all had biochemical and MRI evidence of myositis. Two of these three patients had evidence of weakness at some point in the course of the disease.

Table 1. Clinical features of juvenile dermatomyositis at onset and accrued throughout course of illness
 At presentationa (%)Over disease coursea (%)
  1. a

    Denominator denotes number of patients in whom feature was adequately documented.

Lethargy45/45 (100)47/47 (100)
Weakness54/57 (95)56/57 (98)
Gottron's papules51/56 (91)53/57 (93)
Myalgia/arthralgia46/51 (90) 
Gower's sign40/45 (89)43/49 (88)
Malar rash30/38 (79)36/51 (71)
Heliotrope rash36/49 (73)38/55 (69)
Nailfold changes26/38 (68)38/49 (78)
Weight loss/anorexia25/41 (61)30/48 (65)
Contractures17/29 (59)28/45 (62)
Poikiloderma8/15 (53)11/26 (42)
Dysphonia14/31 (45) 
Mouth ulcers11/29 (38)20/38 (53)
Arthritis15/43 (35)18/49 (37)
Fever16/45 (36) 
Dysphagia11/44 (25) 
Dysphonia or dysphagia 21/49 (43)
Skin ulceration3/13 (23)4/17 (24)
Lymphadenopathy8/35 (23)10/39 (26)
Gastrointestinal involvement6/38 (16)8/40 (20)
Pulmonary involvement4/34 (12)6/36 (17)
Hepatomegaly5/41 (12)6/46 (13)
Calcinosis0/13 (0)8/45 (18)

All patients were documented to have Gottron's papules and/or heliotrope rash at disease onset. Individually, Gottron's papules were seen in 91% (51/56) and heliotrope rash in 73% (36/49). Nailfold capillaroscopy abnormalities were reported in 26 of 38 patients (68%).

Calcinosis was not present in any patient at diagnosis (0/13); however, 18% (8/45) of patients with JDM had calcinosis documented during the course of the disease. Forty-four percent of chronic course patients (7/16) developed calcinosis compared with 4% of monophasic patients (1/21). No patient with polyphasic disease developed calcinosis.

Dysphonia was documented in 14 patients and dysphagia in 11 patients at time of diagnosis. Throughout the course of the illness, 21 of 49 patients (43%) in whom there was adequate documentation had dysphonia and/or dysphagia. At presentation, arthritis was reported in 15 of 43 patients (35%) and contractures in 17 of 29 (59%). Of those patients with contractures at onset, only five (29%) also had arthritis.


Table 2 outlines the results of common investigations performed in the cohort. CK was the most frequently ordered muscle enzyme investigation (100% of patients) and was abnormal 65% of the time (37/57). Twenty patients had normal CK; four of these had no other enzyme measured and 16 had at least one other enzyme and this was abnormal in all cases. Aldolase was measured in only 10 patients and was abnormal in all. When measured, lactate dehydrogenase (LDH), aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were abnormal 92% (23/25), 88% (29/33) and 58% (29/33) of the time, respectively. Two or more muscle enzymes were elevated in 65% of patients (37/57). Four patients (with only CK measured) had no abnormality in muscle enzymes. All four demonstrated clinical weakness and supportive evidence of myositis with abnormal MRI, EMG or muscle biopsy. Erythrocyte sedimentation rate (ESR) was elevated in 84% (46/55) of patients.

Table 2. Results of investigations at diagnosis in descending order of frequency of use
 No. of patients with abnormal result at diagnosis (%)
Creatine kinase37/57 (65)
Aspartate aminotransferase29/33 (88)
Alanine aminotransferase19/33 (58)
Lactate dehydrogenase23/25 (92)
Aldolase10/10 (100)
Erythrocyte sedimentation rate46/55 (84)
Electromyogram4/4 (100)
Magnetic resonance imaging28/29 (97)
Muscle biopsy24/29 (83)
Antinuclear antibodies33/52 (63)
Rheumatoid factor1/18 (6)
Extractable nuclear antigen1/29 (3)
Double-stranded DNA0/15 (0)

Muscle biopsy was performed in 29 patients and was abnormal in 83% (24/29). EMG was performed on four patients and was abnormal in all patients. Figure 2 outlines the frequency of use of muscle biopsy, EMG and MRI in the diagnostic work-up of patients over the period studied. MRI was performed on a total of 29 patients and demonstrated signs of myositis in 97% (28/29). One patient with normal MRI had treatment with oral steroids prior to the MRI.

Figure 2.

Use of Electromyogram, muscle biopsy and magnetic resonance imaging in the diagnosis of juvenile dermatomyositis over the period studied.

Antinuclear antibodies (ANA) were tested in 52 patients and titres were abnormal (titre > 1 : 160) in 33 (63%) cases. High titre antibody to extractable nuclear antigen (ENA) was detected in only one patient (1/32, 3%) and was directed toward topoisomerase-I.


Table 3 outlines therapy at diagnosis and throughout the disease course of the cohort. Fifty-one percent (29/57) of patients were treated with steroids alone (oral and/or high-dose pulsed methylprednisolone) at diagnosis, of whom 12 (20%) received this as their only treatment throughout their disease course. High-dose pulsed intravenous steroids were used in a total of 47 (82%) patients. The majority (89%) of those who received intravenous pulsed steroids did so at the time of diagnosis. Steroids were continued for a median of 18 months (range 10.8–128.7).

Table 3. Therapy in patients with juvenile dermatomyositis (n = 57) at diagnosis and throughout disease course
 At diagnosis (%)Throughout course (%)
Oral steroids48 (84)53 (93)
High dose pulse intravenous steroids42 (74)47 (82)
Methotrexate27 (47)36 (63)
Intravenous immunoglobulin4 (7)18 (32)
Cyclosporine2 (4)10 (18)
Cyclophosphamide0 (0)1 (2)

The combination of steroids and a second-line agent was used in 49% (28/57) of patients at diagnosis and 79% (45/57) during the course of their illness. Sixty-three percent of patients (36/57) were treated with methotrexate (MTX) at some point in the illness and of these, 75% were commenced at diagnosis. Only 14% (4/29) of patients diagnosed prior to 2000 were managed with disease-modifying anti-rheumatic drugs (DMARDs) at diagnosis compared with 86% (24/28) of those managed after 2000 (Fig. 3).

Figure 3.

Use of combination disease-modifying anti-rheumatic drugs (DMARDs) and steroid therapy at diagnosis in juvenile dermatomyositis.

Disease course

Disease course was determined in 45 (79%) patients. The remaining 12 patients had less than 36 months follow-up. The disease was monophasic in 46.7% (21/45), polyphasic in 17.7% (8/45) and chronic in 35.5% (16/45). For monophasic, polyphasic and chronic course, the median time to first remission was 15.7, 22 and 57.7 months, respectively. For the entire cohort, the median time to first remission was 22.3 months.

Nine patients relapsed following a period of remission, eight with polyphasic disease and one with chronic disease. The median time to relapse for patients with polyphasic disease was 11 months (range: 8.0–20.8).


Our cohort demonstrates similar epidemiological and clinical characteristics to those reported from centres in North America, South America, Japan and Europe.[1, 2, 4, 9-14] We have confirmed that female predominance, pre-pubertal onset and a significant duration of symptoms prior to diagnosis, are common epidemiological features of this disease. We have also shown that there are a broad range of clinical features in addition to skin rash and muscle weakness which comprise the clinical syndrome of JDM.

Not unexpectedly the most frequently observed clinical features at diagnosis were weakness and typical rash. Also common were myalgia, arthralgia and nailfold changes. The frequencies of these features are comparable to other studies.[1, 2, 9-11, 13-15] Calcinosis was not seen in any of our patients at diagnosis and was observed in only 18% of cases throughout the disease course. This is lower than reported rates at other centres where rates of calcinosis of up to 40% have been reported.[9-12, 14, 16, 17] The reason for the lower rates of calcinosis in the present study is unclear. It has been postulated that a longer duration of symptoms prior to diagnosis increases the risk of developing calcinosis.[10] However, the time to diagnosis in our cohort was similar to those in papers reporting higher rates of this complication. It is possible that the lower rates of calcinosis in our cohort reflect the more aggressive approach to treatment in recent years.

Muscle enzymes have been reported in the literature to be abnormal in up to 90% of patients with JDM[10]; however, individual enzymes appear to be abnormal at lower rates.[2, 10] CK was found to be a relatively insensitive marker of muscle inflammation in our cohort, being abnormal in just 65% of patients at diagnosis. Other muscle enzymes, such as AST and particularly LDH, were more frequently abnormal, an observation that has been confirmed by others.[2, 10] Our study results support the approach of testing multiple muscle enzymes in the investigation of patients with suspected JDM to increase the sensitivity of these tests for detection of myositis.

The availability of MRI has seen a dramatic decline in the use of EMG and muscle biopsy in the diagnosis of JDM at our centre. This despite the fact that they comprise an important part of the Bohan and Peter criteria, which remain the only validated tool for the diagnosis of JDM. Muscle biopsy was performed in only half of the patients in our cohort and in only 14% of those diagnosed after 2000. EMG was performed in only 7% of patients and in none since 1994. Conversely, MRI was used in the vast majority of patients diagnosed after 2000 and, after muscle enzymes, has become the most frequently used investigation in the diagnosis of JDM. These trends in the diagnostic workup of JDM have been found at other centres[2] and raise the question of whether new criteria for the diagnosis of JDM reflecting modern investigative modalities should be considered.

The treatment of JDM has changed significantly over the last 20 years; the aggressive use of corticosteroids and early initiation of second-line immunosuppressive therapy have become routine practice in many centres, based on data suggesting improved functional outcome and decreased rates of complications, including calcinosis.[10, 12, 18-22] This is reflected in changes in the treatment approach at our centre over the period examined. Prior to 2000, only 14% of our patients were managed with both steroids and a DMARD at diagnosis compared to 86% of those patients managed after 2000. It is difficult to draw conclusions regarding the outcomes of different treatment modalities given the range of regimens in our cohort.

The findings of this study should be considered in light of a number of possible limitations. This study was a small retrospective review and there was incomplete documentation of findings, especially with respect to the absence of less common clinical features. In addition, the data collected on many clinical features was subjective and therefore reliant on individual clinician acumen. The search technique may have introduced a selection bias as only patients admitted to hospital were identified. Patients managed solely as outpatients would not have been included, potentially over-estimating the severity and treatment requirements of the disease.

This Australian cohort of patients with JDM revealed characteristics similar to previously described cohorts and adds to the global data of this rare disease. This study demonstrates a change in diagnostic approach over time, emphasising the current role of MRI, and also outlines the importance for measurement of multiple muscle enzymes. The early use of DMARDs has become common. Although the outcome for children with JDM has improved, it remains a disease requiring long-term care with a largely unpredictable course.

Conflict of Interest

The authors declare. No conflict of interest