Article Text

Download PDFPDF

Extended report
Pulmonary outcome in juvenile dermatomyositis: a case-control study
  1. Helga Sanner1,2,
  2. Trond Mogens Aaløkken3,
  3. Jan Tore Gran1,2,
  4. Ivar Sjaastad1,4,5,
  5. Bjørn Johansen1,6,
  6. Berit Flatø1,2
  1. 1Institute of Clinical Medicine, University of Oslo, Oslo, Norway
  2. 2Department of Rheumatology, Oslo University Hospital, Rikshospitalet, Oslo, Norway
  3. 3Department of Radiology, Oslo University Hospital, Rikshospitalet, Oslo, Norway
  4. 4Institute for Experimental Medical Research, Oslo University Hospital, Oslo, Norway
  5. 5Department of Cardiology, Oslo University Hospital, Oslo, Norway
  6. 6Department of Respiratory Medicine, Oslo University Hospital, Rikshospitalet, Oslo, Norway
  1. Correspondence to Dr Helga Sanner, Department of Rheumatology, Oslo University Hospital, Rikshospitalet, 0027 Oslo, Norway; helga.sanner{at}


Objectives To compare pulmonary function in patients with juvenile dermatomyositis (JDM) with that of matched controls; and to examine associations between pulmonary function impairment, high-resolution CT (HRCT) abnormalities and other disease variables in patients with JDM.

Methods A total of 59 patients with JDM clinically examined a median 16.8 years (range 2–38 years) after disease onset were compared with 59 age-matched and sex-matched controls. Pulmonary function tests included spirometry, diffusing capacity for carbon monoxide (DLCO) and body plethysmography. In patients with JDM, HRCT scans were performed and cumulative organ damage and patient-reported health status assessed.

Results Patients with JDM had lower total lung capacity (TLC) and DLCO compared to controls (p=0.003 and <0.001, respectively). A low TLC was found in 26% of patients versus 9% of controls (p=0.026), and a low DLCO in 49% of patients versus 9% of controls (p<0.001). HRCT abnormalities were found in 37% of patients, and included interstitial lung disease (ILD) (14%), chest wall calcinosis (14%) and airway disease (15%). Three patients were diagnosed as having ILD prior to the follow-up visit. A low TLC was more often found in patients with than without abnormal HRCT (50% vs 12%, p=0.002). HRCT abnormality correlated with cumulative organ damage (rs=0.346, p=0.008) and patient-reported health status at follow-up (p<0.005).

Conclusions Patients with JDM had smaller lung volumes than controls; a restrictive ventilatory defect was found in 26% and HRCT abnormality in 37% of the patients, and these findings were associated. Although mostly subclinical, the relatively high frequency of pulmonary involvement highlights the systemic nature of JDM.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Juvenile dermatomyositis (JDM) is a chronic vasculopathic disease of childhood. In adult polymyositis (PM) and dermatomyositis (DM), interstitial lung disease (ILD) is a frequent complication being associated with high morbidity and mortality,1 while the data on pulmonary involvement in JDM is sparse. An asymptomatic impairment in pulmonary function was found in 5/12 patients with JDM in a longitudinal study; alveolar volume (VA) and diffusion lung capacity of carbon monoxide (DLCO) being the most frequently reduced variables.2 Furthermore, 5/10 patients with severe JDM were diagnosed as having ILD, confirmed by high-resolution CT (HRCT) in suspected cases.3 Serious or fatal pulmonary complications in JDM have been described in case reports.4 5 Also, pulmonary involvement in JDM has been described in studies not primarily addressing this aspect; asymptomatic significant restrictive impairment being found in 14/17 patients6 and ILD in 7% to 19%.7,,9

These small series suggest that restrictive ventilatory defects (defined by reduced lung volumes) are common in JDM, and might be caused by reduced compliance of the lungs (for example, ILD) or reduced compliance of the chest wall due to muscle weakness. Also, calcinosis may be confined to the chest wall10 and thus cause restriction,11 but the effect on pulmonary function has never been systematically investigated. The preferred method to diagnose restrictive ventilatory defects is body plethysmographic assessed total lung capacity (TLC).12 HRCT is a sensitive and non-invasive technique for detecting pulmonary involvement, such as ILD.13 However, no studies have addressed pulmonary outcome using HRCT and pulmonary function tests (PFT), including body plethysmography, in an unselected JDM cohort, and no controlled studies of pulmonary function in JDM are available.

Thus, the aims of the present study were to compare lung function in a JDM cohort with controls from the general population, and to determine the prevalence of abnormal HRCT and pulmonary function impairment in patients. Secondary aims were to investigate whether PFT and HRCT findings in the patients correlated with each other, and with patient's characteristics and disease variables assessed at the first year post diagnosis and at follow-up.

Patients and methods

Patients and controls

Inclusion criteria were: a probable/definitive diagnosis of DM according to the Bohan and Peter criteria,14 disease onset <18 years, minimum 24 months of disease duration and age ≥6 years at inclusion. A retrospective inception cohort of patients from Norway diagnosed as having JDM between January 1970 and June 2006 were identified by electronic and manual search in the archives from this time period.15 In all, 66 patients fulfilled the inclusion criteria, 4 of whom were deceased; the remaining 62 could all be tracked (via the Norwegian Population Register) and 59 (95%) participated in the study.

Controls from the general population (one sex-matched and age-matched control per patient) as previously described16 were randomly selected from the National Population Register; exclusion criteria were lung or heart diseases (except for mild asthma), mobility problems, inflammatory rheumatic disease and other autoimmune diseases treated with immunosuppressive medication.

Informed consent was obtained from all patients and controls (and their parents if aged <16 years), according to the Declaration of Helsinki. The study was approved by the Regional Ethics Committee for Medical Research.

Data collection and clinical measures

All patients and controls were examined during a 1–2 day programme at Oslo University Hospital in the period September 2005 to May 2009. This included clinical examination by a single doctor (HS), PFT and blood samples. Smoking habits were recorded.

In patients, HRCT was performed and anti-nuclear antibodies (ANA) analysed. Disease activity was measured by Disease Activity Score for JDM (DAS)17 and cumulative organ damage by Myositis Damage Index (MDI).18 The DAS and MDI were also scored retrospectively from the first year post diagnosis (as previously described).15 The Health Assessment Questionnaire (HAQ)19 and the Child HAQ20 were used to measure physical function in patients aged ≥18 years (n=39) and <18 years (n=20), respectively. Physical health was measured by the Short Form-36 (SF-36) physical component summary (PCS).21 Information of immunosuppressive medication was obtained from the medical records.

Pulmonary function tests

PFT included spirometry, measurement of gas diffusion and body plethysmography. All measurements were performed on a computerised Vmax Pulmonary Function Unit (Viasys, Santa Ana, California, USA). Spirometric variables (measured in triplicate) included the forced vital capacity (FVC) and the forced expiratory volume in 1 s (FEV1). Gas diffusion variables (measured in duplicate) included DLCO, VA and the transfer coefficient (DLCO/VA). All diffusion variables were corrected for haemoglobin. The body plethysmographic variables (measured in duplicate) included TLC and vital capacity (VC) and were carried out in 55 patients and 54 controls (not performed in individuals aged <9 years, n=3; and in 2 controls and 1 patient due to technical/practical problems). All lung function measurements were performed according to published guidelines.22,,24 The pulmonary function values were expressed either in absolute terms or as percentage of predicted. Predicted values were derived from reference equations, separate for each gender, with age and height as predictor variables.25 26 Low TLC, DLCO and DLCO/VA were defined as less than the fifth percentile of the predicted values.12


HRCT was carried out in 57 patients using a LightSpeed 16 scanner (GE Healthcare, Milwaukee, Wisconsin, USA). Thin section CT images were obtained in the supine position (during deep inspiration). In order to minimise the radiation dose, 120 kV, 40–200 mA, 0.8 s rotation time with 1.25 mm section thickness at 10 mm intervals were used, and tube current setting was adjusted to each patient's age and weight. Supplementary expiratory scans were obtained in 48 patients.

The images were reviewed by an experienced radiologist blinded to clinical information. The presence, extent and distribution for established CT criteria of ILD and airway disease27 were evaluated and included ground-glass opacity, airspace consolidation, reticular pattern, interlobular septal thickening, cysts, traction bronchiectasis/bronchiolectasies, micronodules, pleural irregularity and air trapping. The distribution of changes was reviewed in four zones: (A) above the aortic arch, (B) between the aortic arch and the level of carina, (C) between the level of carina and the level of the inferior pulmonary veins and (D) below the inferior pulmonary veins.

The extent of ground-glass opacity and reticular pattern were scored based on the percentage of lung parenchyma that showed evidence of an abnormality (1=1% to 5%, 2=6% to 10%, 3=11% to 20%; 4=21% to 50%; and 5=>50%). The severity of bronchiectasis was scored 0–4 (1=bronchial wall thickening without distinct ectasis, 2=mild, 3=moderate and 4=severe bronchiectasis). Micronodules were scored 0–3 (1=mild, 2=moderate, 3=severe). For all four parameters, the scores in the four localisations (A–D) were summed and divided by four, in order to calculate an extent/severity score.

Calcinosis in the chest wall (subcutis, fasciae and muscle) was registered, and the number of foci was counted. The extent of calcinosis was scored: 1=1–5 spots, 2=6–10 spots, 3=>10 spots. Pleural thickening was evaluated in all sections and scored based on the % of pleural area involved; 1=1% to 5%, 2=6% to 10%, 3=11% to 20% and 4=>20%.

HRCT-detected ILD was defined as reticular pattern (with/without traction bronchiectasis) and/or ground glass opacity, whereas HRCT-detected airway disease was defined as bronchiectasis, and/or air trapping and/or micronodules.

Statistical analysis

Differences between patients and matched controls were tested by the paired sample t test (normally distributed continuous variables), Wilcoxon's test (continuous not normally distributed or ordinal variables) or McNemar's test (binary data). Differences between patient groups were tested by the independent sample t test, the Mann–Whitney U test or χ2 test as appropriate. Correlations were determined by the Spearman correlation coefficient (rsp). All tests were two tailed. SPSS V.16.0 (SPSS, Chicago, Illinois, USA) was used for statistical analyses.


Characteristics and clinical lung involvement

The age difference between patients and controls was median −0.4 months (range −15.4 to 11.6 months), whereas the absolute age difference was <6 months in 51 pairs (87%). No differences in characteristics were found between patients and controls (table 1). None of the controls had lung symptoms. Three patients had been diagnosed as having ILD prior to the follow-up visit. One girl was diagnosed as having JDM and anti-Jo1 positive, biopsy verified fibrosing alveolitis at the age of 14 years; one boy with JDM from the age of 2 developed symptomatic ILD 13 years later. Also, one boy with JDM from the age of 16 developed ILD complicated with mediastinal emphysema 2 months post diagnosis. Furthermore, one girl, diagnosed as having JDM at the age of 15, developed mediastinal emphysema without evidence of ILD, 6 months post diagnosis. All four of these patients had arthritis during their disease course; both patients with mediastinal emphysema had a history of Raynaud's phenomenon and skin ulcers.

Table 1

Characteristics and lung symptoms in patients with juvenile dermatomyositis (JDM) and controls

Pulmonary function in patients and controls

A low TLC was found in 14/55 (26%) patients and 5/54 (9%) controls; (p=0.026). A low DLCO was found in 29/59 (49%) patients and 5/59 (8%) controls (p<0.001). None of the patients or controls had a decreased DLCO/VA (transfer coefficient). FVC, FEV1, VC, TLC and DLCO were lower in patients than controls (all p<0.003) (table 2). These variables were also lower in the patients, after excluding subjects with known ILD and their respective controls form the analyses (data not shown). Selected PFTs (% of predicted values) are shown in figure 1.

Figure 1

A–D. Pulmonary function tests in patients and controls. Data presented are the percentage of predicted; predicted values are taken from reference equations based on sex, height and age. Diffusion variables are also corrected for haemoglobin. Data are presented as box plots, were the boxes represent the 25th–75th percentile, the line within the boxes represent the median; n=59 for patients and controls unless otherwise stated. B. n=55 for patients and 54 for controls. DLCO, diffusing capacity for carbon monoxide; DLCO/VA, DLCO per litre alveolar volume (transfer coefficient); FVC, forced vital capacity; TLC, total lung capacity.

Table 2

Pulmonary function in patients with juvenile dermatomyositis (JDM) and controls assessed at follow-up

HRCT findings in patients

An abnormal HRCT of the thorax (any HRCT finding) was found in 21/57 (37%) patients (table 3). Changes compatible with ILD were found in 8/57 (14%) patients and fine intralobular fibrosis was most common. One patient (anti-Jo1 positive) had macrocystic reticular pattern (honeycombing) in addition to dorsal ground glass opacity. Three of the patients with HRCT-detected ILD also had traction bronchiectasis. HRCT-detected airway disease was found in 9/57 (15%) patients. Of the five patients with bronchiectasis, four either had asthma (n=2) or were currently/previously daily smokers (n=3). HRCT-detected calcinosis was found in 8/57 (14%) patients, and was localised in the fasciae in 7 patients, intramuscularly in 2 and subcutaneously in 2 (in 3 patients a mixture was seen). Two patients had HRCT-detected ILD and airway disease (air trapping), one patient had ILD and calcinosis, and one had calcinosis and airway disease.

Table 3

HRCT findings in patients with juvenile dermatomyositis (JDM) at follow-up

Associations between pulmonary outcome, patient characteristics and disease variables assessed at follow-up

A low TLC was found in 10/20 (50%) patients with versus 4/33 (12%) patients without an abnormal HRCT (p=0.002), and in 4/7 (57%) with versus 10/46 (22%) without HRCT-detected calcinosis in muscle and/or fascia (p=0.048). TLC% of predicted correlated with HRCT abnormality including HRCT-detected airway disease, whereas DLCO% of predicted did not correlate with any HRCT variables (table 4). Most patients with a low DLCO did not have a low TLC (19/27, 70%) and HRCT-detected ILD was not found in any of these patients. No significant correlation was found between DLCO% and TLC% (rs=0.218, p=0.110).

Table 4

Correlations between HRCT findings and other disease variables in patients with juvenile dermatomyositis (JDM)

HRCT abnormality correlated with total MDI, HAQ/child-HAQ and SF-36 PCS at follow-up (table 4). HRCT findings were not associated with disease activity (DAS), muscle strength, ANA or muscle enzymes (data not shown). Neither a low TLC nor a low DLCO, were associated with any of these disease variables assessed at follow-up. Dyspnoea at exertion was reported in 3/8 (38%) with versus 4/49 (8%) without HRCT-detected ILD (p=0.050). Use of methotrexate during disease course, (30/59, 51%) was not associated with HRCT-detected ILD (data not shown).

Smoking habits (currently or previously smoking) were not significantly associated with low TLC, low DLCO or any HRCT findings (data not shown). No differences in the prevalence of PFT impairment (low DLCO or low TLC) or HRCT findings were found when comparing patients whose diagnosis was made before and after 1990 (figure 2).

Figure 2

Frequency (% of patients with juvenile dermatomyositis (JDM)) with an unfavourable pulmonary outcome, by all patients, patients diagnosed before and patients diagnosed after 1990. No difference was found between patients diagnosed before and after 1990 for any abnormalities. PFT abnormalities=low TLC and/or low DLCO. DLCO, diffusing capacity for carbon monoxide; HRCT, high-resolution CT; ILD, interstitial lung disease; PFT, pulmonary function tests; TLC, total lung capacity.

Association between pulmonary outcome and early disease variables

Neither HRCT abnormality nor low TLC or low DLCO, correlated with disease activity, muscle enzymes or treatment variables assessed at diagnosis or 6 and 12 months post diagnosis, or with age at diagnosis/duration of untreated disease (data not shown). HRCT-detected ILD correlated with MDI total at 6 and 12 months post diagnosis (rs=0.325, p=0.014 and rs=0.274, p=0.039).


At a median 16.8 years after disease onset, patients with JDM had lower TLC and DLCO compared to matched controls. A low TLC was found in approximately 25% and a low DLCO in approximately 50% of the patients. HRCT abnormalities were detected in 37% of the patients, and included ILD, chest wall calcinosis and airway disease. HRCT abnormality was associated with a low TLC and also correlated with cumulative organ damage and patient-reported health status at follow-up, but not with early disease activity. To our knowledge, this is the first controlled study to examine pulmonary function in JDM and also to address pulmonary outcome in an unselected JDM cohort, using HRCT and body plethysmography.

We have previously described the representativeness of our patient cohort.15 The frequency of dyspnoea in our patients is in accordance with previous reports,28 and the frequency of patients with known ILD (5%) is comparable with that in a registry based cohort.9 The randomly selection of our controls from the National Population Register, is a strength of our study. Also, height, smoking habits and reported asthma at follow-up did not differ between the patients and controls.

The patients had lower FVC, FEV1, DLCO, VC and TLC than the controls. The reduction in TLC means that patients had smaller lung volumes, which explains the reduced FVC and FEV1 seen in the patient group. Also, the lower DLCO might be a consequence of reduced lung volumes. The transfer coefficient (DLCO/VA) was nominally lower, albeit not reaching statistical significance, further indicating a volume problem in the patients. The preferred method to diagnose restrictive pulmonary impairment is body plethysmographic assessed TLC.24 Using established cut-off values,12 26% of our patients had restrictive ventilatory defects, which is in accordance with previous reports,2 6 and may be a consequence of pulmonary diseases such as ILD, or be caused by extrapulmonary conditions such as respiratory muscle weakness, calcinosis in the chest wall or pleural disease.

HRCT abnormality was found in 37% of our patients; ILD, chest wall calcinosis and airway disease being equally frequent (approximately 15%). Compared to adult DM/PM, ILD was found less frequently in our cohort.29 In the present report, a low TLC was associated with HRCT abnormalities, and also with HRCT-detected calcinosis and airway disease, but not with HRCT-detected ILD. The association between chest wall calcinosis and restriction is not surprising, since calcium deposits might lead to respiratory muscle impairment11; however, the association between a low TLC and airway disease (eg, bronchiectasis) is more difficult to explain and might be due to a type 1 error. Airway involvement is rare in adult PM/DM,30 but has been described in 2/12 patients with JDM.2 ILD is known to be associated with restriction,31 but in our study ILD in most patients was limited to <10% of the lung volumes, which explains the lack of association between ILD and TLC. We might also have been underpowered to show such an association. HRCT-detected ILD is generally not found in the normal population.32 Also, it was not found in any of 60 patients with childhood systemic lupus erythematosus (mean age 28 years) from our centre, and HRCT abnormalities were found in only 5/60 (8%).33

Almost half of our patients had a low DLCO, but none of them had a low transfer coefficient (DLCO/VA). Even though often reported, DLCO/VA might not be helpful in the differential diagnosis of volume restriction.23 34 Of the patients with a low DLCO, 70% had no restriction and none of these patients had HRCT changes consistent with ILD. Since normal lung volumes associated with impaired diffusion may suggest pulmonary disease,12 this finding may indicate the presence of ultrastructural changes in the alveolar membrane35 or pulmonary vascular disease.36

HRCT abnormalities correlated with cumulative organ damage and poorer patient-reported health status. Also, a borderline significant association was found between HRCT-detected ILD and dyspnoea on exertion. Taken together, we believe this supports the clinical relevance of our findings. Even though approximately 75% of the patients either had impaired diffusion, restriction or HRCT abnormality at follow-up, most patients did not report lung symptoms. However, one can speculate that the lung symptoms may have been masked by restricted functions in other organ systems. For example, if patients are not able to complete physical exercise due to muscle weakness, they will not experience shortness of breath even when they have reduced lung function. Thus, some patients do not report dyspnea even though they may have reduced lung function. Follow-up studies are needed in order to investigate whether more of our patients with detectable HRCT abnormalities and PFT impairments will develop clinical manifest pulmonary disease in the future.

Studying the same cohort, we have previously shown that sustained early disease activity predicts cumulative organ damage and an unfavourable muscle outcome.15 16 In the present study, no association was found between early disease activity and pulmonary outcome. It may be possible, that the disease process affecting the pulmonary system differs from the processes in other organs. However, HRCT abnormality correlated weakly with a high cumulative prednisolone dose. Since there is no evidence that prednisolone can cause the described HRCT findings, it is possible that a high prednisolone does merely reflect longstanding active disease,

No statistical significant difference in pulmonary outcome was found between patients diagnosed before rather than after 1990, even though the first group have been treated less aggressively.15 However, there was a trend towards more restriction and HRCT abnormalities (especially HRCT-detected ILD) in the earliest diagnosed patients. Our study might have been underpowered to detect such differences. However, it is believed that ILD can develop in the chronic phases of the disease,2 which is also illustrated by the development of symptomatic ILD in one of our patients 12 years post diagnosis. Some of our latest diagnosed patients may still develop ILD. Thus, we are limited by the lack of longitudinal assessment of the outcome measures. We are also limited by the retrospective assessment of early disease variables.

In conclusion, patients with JDM had smaller lung volumes than the controls, and restriction, HRCT abnormalities and impaired diffusion were found in approximately 25% to 50% of the patients. Although most patients did not report lung symptoms, HRCT abnormalities correlated with restriction, cumulative organ damage and patients reported health status, indicating a clinical relevance of our findings. Also, our results highlight the systemic nature of JDM.


We thank Are Hugo Pripp for statistical advice and Wenke Lunde and the other staff at the Pulmonary Function Laboratory for performance of the pulmonary function tests.



  • Funding The project received financial support form Dr Olga Imerslunds foundation, Oslo, Norway

  • Competing interests None.

  • Ethics approval This study was carried out in compliance with the declaration of Helsinki and was approved by the Regional Ethics Committee, (Helse Sør-Øst).

  • Provenance and peer review Not commissioned; externally peer reviewed.