The effectiveness of treating juvenile dermatomyositis with methotrexate and aggressively tapered corticosteroids




Childhood dermatomyositis (DM) is often a chronic disease, lasting many years. It has traditionally been treated with long-term corticosteroid therapy; side effects are often seen. For more than a decade, methotrexate (MTX) has been safely used for the treatment of juvenile arthritis. Here, we report use of MTX as first-line therapy for DM, along with aggressively tapered corticosteroids, in an attempt to reduce treatment-related side effects.


We studied an inception cohort of 31 children with DM who were rigorously followed up in our myositis clinic, and compared them with a control group of 22 patients with incident cases of juvenile DM who received treatment just before we instituted a policy of first-line therapy with MTX. The mean starting dosage of MTX in the study group was 15 mg/m2/week.


Both groups had similar improvement in strength and physical function; however, the median time during which patients in the study group received corticosteroids was 10 months, compared with 27 months for controls (P < 0.0001). As a result, the cumulative prednisone dose in the study group was approximately half that in the control group (7,574 mg versus 15,152 mg; P = 0.0006). The study group had greater height velocity during the first year of treatment and a smaller increase in the body mass index over the first 2 years. In the control group, the relative risk of cataracts developing was 1.95 (95% confidence interval 1.05–4.17). Side effects of MTX were rarely observed.


Use of MTX in conjunction with an aggressively tapered course of prednisone may be as effective as traditional long-term corticosteroid therapy for children with DM, while decreasing the cumulative dose of corticosteroids.

Juvenile dermatomyositis (DM) is a chronic multisystem disease of presumed autoimmune origin that primarily affects skin and muscle. In western populations, the incidence of juvenile DM is ∼2–3 per 1 million children per year (1–3). Although the outcomes of juvenile DM have improved over the last few decades, the disease is still associated with significant morbidity in a proportion of patients (4, 5). For many patients, juvenile DM follows a chronic course. According to a recent retrospective multicenter Canadian study of children with juvenile DM, 35% of the children (23 of 65) were still receiving medication for persistent disease activity after a median followup of 7 years (5).

No prospective randomized study has shown the efficacy of any immunosuppressive agent for the treatment of juvenile DM. The mainstay of therapy for juvenile DM has been long-term corticosteroid treatment, which, although efficacious, is associated with significant side effects. The standard approach to treatment of juvenile DM has been use of high doses of oral corticosteroids, with slow tapering over at least 2 years (6). The major limitations of this treatment are side effects and, in some patients, incomplete response.

The major side effects of corticosteroids include growth delay, osteoporosis, avascular necrosis, cataracts, and hypertension. In 2 studies (7, 8), side effects of corticosteroids were observed in 32–41% of adult patients with DM or polymyositis. Although the relationship between cumulative doses of corticosteroids and these side effects remains unclear, in general the side effects of corticosteroids are thought to be related to the dose and duration of therapy. It is increasingly recognized that even prolonged low-dose therapy may cause some of these side effects (9).

Methotrexate (MTX) is an immunosuppressive agent that has been used successfully in a wide variety of pediatric rheumatic diseases (10, 11). Experience gained from the use of MTX in childhood arthritis has shown that low-dose therapy is safe, with little reported long-term toxicity. In small, uncontrolled, retrospective studies (12–16), MTX was suggested to be beneficial as adjunctive therapy in patients with juvenile DM refractory to corticosteroids or in patients dependent on corticosteroids (12–16). Recently, Fisler et al conducted a review of the medical records of 35 children with juvenile DM, 23 of whom were started on a regimen of MTX after failing to respond within 6 weeks to corticosteroids. This cohort was considered to have had a good response to treatment and a decreased incidence of calcinosis (17). At followup, 5 patients (14%) had mild calcinosis; this rate is lower than the rate of ∼33% observed in other studies (18).

As a result of these studies, there has been interest in starting MTX earlier in patients with disease refractory to corticosteroid treatment, similar to the early use of MTX or azathioprine in adult DM. However, corticosteroids remain the initial treatment for the majority of patients with juvenile DM. For example, Wulffraat et al surveyed 50 pediatric rheumatology centers in 30 countries representing the Paediatric Rheumatology International Trials Organization. They found that in all centers corticosteroids were used initially, and that only 17 centers considered using MTX adjunctively as a first-line agent (19). There are no controlled studies of MTX as a first-line agent, with or without concomitant steroid therapy, in the treatment of patients with juvenile DM.

Our research question was as follows: does treatment of juvenile DM with MTX and an aggressively tapered course of corticosteroids result in a decreased cumulative dose of corticosteroids, while not compromising treatment efficacy, when compared with the standard regimen in which corticosteroids are tapered more slowly? We hypothesized that our approach of using MTX along with an aggressively tapered course of corticosteroids is as effective as the traditional regimen, results in a lower cumulative dose of corticosteroids and fewer corticosteroid-related adverse effects, and is safe in terms of MTX side effects.

We present the results of our cohort study, in which patients treated with MTX in combination with a course of aggressively tapered corticosteroids were compared with historical controls from our institution who were treated according to the traditional approach of a lengthy course of corticosteroids.


Patient population and study design.

We performed a retrospective cohort study of 31 consecutive patients with incident cases of juvenile DM who were seen in our clinic since June 1997. We compared these patients with 22 consecutive control patients with incident cases of juvenile DM who were seen in our clinic during the preceding period, January 1993 to May 1997. The start date of January 1993 was chosen because at that time we began using a more comprehensive followup form at the clinic. All 53 patients underwent the same protocol-based followup in the myositis clinic at the Hospital for Sick Children (Sick Kids). The study was approved by the Research Ethics Board at Sick Kids.

Inclusion and exclusion criteria.

All children in whom probable or definite juvenile DM was diagnosed at Sick Kids were included (20, 21). Patients in whom a diagnosis was made and for whom therapy began before referral to Sick Kids and those children who had myositis as part of an overlap syndrome were excluded (Figure 1).

Figure 1.

Disposition of children with juvenile dermatomyositis (JDM) seen at The Hospital for Sick Children (HSC) during the study period. JPM = juvenile polymyositis.

Sick Kids myositis clinic.

Since 1991, all patients with juvenile DM or other idiopathic inflammatory myopathies who were seen at Sick Kids have been followed up in a separate subspecialty clinic. Staff members attending the juvenile DM clinic include a rheumatologist, a nurse specialist, a dedicated physiotherapist (and more recently a physiotherapy practitioner), a dedicated nutritionist, and, most recently, a neurologist specializing in neuromuscular disorders. Patients are attended to by the house staff rotating through the clinic, who are supervised by the faculty. All patients are assessed and treated according to a defined protocol that has been in place since the inception of the clinic. Collaborating members of the divisions of dermatology, neurology, and ophthalmology see all of the patients early in the course of their disease. All clinic and laboratory data are prospectively collected and stored in a dedicated database.


The study group received MTX (10–20 mg/m2/week [maximum dosage 25 mg/week]) and prednisone (2 mg/kg/day [maximum dosage 75 mg/day]). Prednisone initially was administered in 3 divided doses and then was consolidated, over the first 6 weeks, into a single daily dose. At 6 weeks, if the level of muscle enzymes in serum was reduced compared with that at the time of diagnosis and the patient had improved clinically, the prednisone dose was tapered by ∼10% every 2 weeks. Weaning patients off prednisone was done according to protocol; however, patients were not weaned if disease was not well controlled. If there was a flare of symptoms, the dose of prednisone was increased until control was achieved. Subjects were evaluated every 2 weeks initially; the frequency of visits then decreased to every 4 weeks, then every 6–12 weeks over the first year. Initially, MTX was administered orally; if required, the route of administration was changed to subcutaneous. Children were given MTX subcutaneously if nausea and/or vomiting occurred as a side effect or if a dosage in excess of 15 mg/week was required.

The control group received prednisone at a dosage of 2 mg/kg/day. Prednisone initially was administered in divided doses and then was consolidated, over the first 6 weeks, into a single daily dose. The patients were then weaned off prednisone, with tapering at a rate of ∼10% every month, taking into account the clinical response.

In children presenting with severe weakness, dysphagia, dysphonia, or vasculitic skin ulcers (ulcerative juvenile DM), additional medications including cyclophosphamide, cyclosporine, pulse intravenous methylprednisolone (30 mg/kg, to a maximum dosage of 1,000 mg/day for 3 days), and/or intravenous immunoglobulin (IVIG) were added at the onset of treatment, at the discretion of the attending rheumatologist. Hydroxychloroquine and IVIG were also used in children in whom skin rash persisted despite other therapy.


All control patients were followed up for 4 years (1,056 patient-months). The study patients were followed up for an average of 34.6 months. Of the 31 study patients, 12 were followed up for 4 years, 2 were followed up for 3–4 years, 12 were followed up for 2–3 years, 2 were followed up for 1–2 years, and 3 were followed up for <1 year (1,073 patient-months).

Assessment and outcome measures.

The primary outcome measure was the time to discontinuation of prednisone, as well as the cumulative corticosteroid dose (including prednisone and intravenous methylprednisolone) and the presence of corticosteroid side effects (see below) at 3, 6, 12, 18, 24, 36, and 48 months of followup.

Efficacy, as assessed by muscle power, the Childhood Health Assessment Questionnaire (C-HAQ; a validated measure of physical disability in juvenile DM) (22, 23), the revised American College of Rheumatology (ACR) functional class (24), presence of skin rash, time to development of calcinosis, and time to flare, was also examined. At each followup visit, the presence of calcinosis was determined clinically, and nailfold capillary microscopy was performed. Because muscle enzyme levels normalize rapidly at the onset of therapy (as indeed they did in both of our groups) and are not well correlated with other measures of disease activity, we did not use muscle enzyme levels for the purposes of this study.

We examined corticosteroid side effects by calculating the change in body mass index (BMI) as a proxy for cushingoid obesity, height velocity, time to development of cataracts, and time to development of spinal fractures. Bone mineral density studies were not routinely performed in our control patients and therefore could not be used to compare the 2 treatment strategies. Significant gastrointestinal (GI) toxicity (defined as toxicity leading to cessation of therapy) was used as a measure of MTX toxicity.

The muscle power of all patients was scored by the physician at each clinic visit, using the Oxford Medical Research Council grading system (25). Additionally, the clinic physiotherapist scored each of the major muscle groups using dynamometry (using a modified sphygmomanometer) (26, 27).

Disease flares were defined as active disease of skin or muscle requiring increased therapy with corticosteroids and/or additional medications. Liver-related enzyme levels (alanine aminotransferase, aspartate aminotransferase) were considered to be abnormal if they were twice the upper limit of normal.

Statistical analysis.

Because both groups were not followed up for the same length of time, we used survival analysis to answer our primary question. The time to event end points (e.g., discontinuation of corticosteroids, time to flare, time to development of cataract) were plotted using the Kaplan-Meier technique, and the 2 treatment groups were compared using the log rank test; relative risks (RRs; relative hazards) were calculated using proportional hazards regression models. Other comparisons between the groups used repeated-measures analysis of variance and the 2-sample t-test. When the assumptions for parametric testing were not met, nonparametric testing was substituted. Because there are no normal standards for children for the muscle dynamometry testing that we used, we calculated the change from baseline for each subject as an absolute percent change (compared with the maximum score of 300). The analyses were carried out using Data Desk statistical software version 6.2 (Data Description, Ithaca, NY) and the JMP statistical package version 5.0.1a (SAS Institute, Cary, NC).



The 2 groups comprised 31 study patients and 22 control patients, respectively. The baseline characteristics of the 2 groups, including age at diagnosis, sex, and ethnicity, were similar (Table 1). Baseline function and strength, as measured by the C-HAQ, median muscle strength in the weakest muscle group, and revised ACR functional class, were also similar in the 2 groups. All patients in both groups had evidence of skin disease at the time of diagnosis of DM.

Table 1. Baseline characteristics of patients in the 2 groups*
CharacteristicStudy patients (n = 31)Control patients (n = 22)P
  • *

    Except where indicated otherwise, values are the number of patients. The determination of definite and probable dermatomyositis was based on the criteria described by Bohan and Peter (20, 21). C-HAQ = Childhood Health Assessment Questionnaire; ACR = American College of Rheumatology.

Mean age at diagnosis, years8.46.70.11
No. boys/no. girls11/206/160.53
Ethnicity  0.76
 East Indian41 
 Native Canadian11 
Definite dermatomyositis2715 
Probable dermatomyositis470.17
Mean body mass index, kg/m216.9816.390.58
Mean weight, kg30.125.50.32
Mean muscle power in weakest muscle group, kg3.63.50.81
Mean C-HAQ score1.251.480.31
Presence of nailfold abnormalities29160.68
Mean ACR functional class2.802.950.56
Median prednisone starting dosage, mg/kg/day1.812.000.01

Cumulative dose of corticosteroid, time to discontinuation of corticosteroids, and corticosteroid side effects.

The median starting dose of corticosteroid was different in the 2 groups (1.81 mg/kg in the study group versus 2.00 mg/kg in controls; P = 0.01). These differences may reflect the fact that the study group was somewhat older and heavier at baseline and therefore more likely to reach the maximum dosage of 75 mg of prednisone per day.

Use of the described protocol resulted in significantly less exposure to prednisone in the study patients than in the control patients. In most study patients it was possible to closely follow the tapering protocol, but there was some variability in the time to full discontinuation of prednisone (Figure 2). The median time to discontinuation of prednisone was 10 months in the study group, compared with 27 months in the control group (P < 0.0001, RR 0.41, 95% confidence interval [95% CI] 0.29–0.57). The average cumulative dose of corticosteroids (expressed as milligrams of prednisone) in the 2 groups, including only those patients in the study group who were followed up for the full 4 years, was 15,152 mg in the control group and 7,574 mg in the study group (P = 0.0006).

Figure 2.

Survival plot showing time to discontinuation of prednisone. The y axis shows the proportion of subjects who were still receiving prednisone at any point in time. The x axis shows the number of months since starting therapy.

The average BMI at the time of diagnosis was similar in the 2 groups (16.98 kg/m2 in the study group versus 16.39 kg/m2 in the control group). At 1 year of followup, the BMI had increased by an average of 4.7 kg/m2 in the control group, compared with an average increase of 2.2 kg/m2 in the study group (F = 14.5, 1 df, P = 0.0004), and over the first 2 years the BMI had increased by 2.8 kg/m2 in the control group and by 2.0 kg/m2 in the study group (F = 6.67, 2 df, P = 0.002). There were too few data to permit a meaningful comparison of the BMI at 3 and 4 years.

The height velocity at 1 year was significantly different in the 2 groups. The control patients had an average height velocity of 2.8 cm/year compared with an average of 4.1 cm/year in the study group (P = 0.048). The difference in height velocity was no longer discernible at 2 years of followup (both groups grew at the same rate from year 1 to year 2).

Cataracts developed in 8 patients in the control group and in 3 patients in the study group. Fortunately, none of these cases has been vision-limiting or has required therapy. The time to development of cataracts was significantly shorter in the control group (RR 1.95, 95% CI 1.05–4.17, P = 0.03). Spinal fractures developed in 2 control patients (at 3 months and 11 months, respectively) and in 1 study patient (at 6 months).

Disease activity.

During followup, changes in disease activity as measured by the C-HAQ, muscle power, rash, and the revised ACR functional class were similar in the 2 groups. By 12 months, the median C-HAQ was 0 (normal function) in both groups, and over the first 2 years there was no difference between the groups in repeated measures of improvement (F = 1.3, 5 df, P = 0.27). Likewise, the improvement in functional class was similar between the 2 groups over the first 2 years of followup (F = 0.72, 5 df, P = 0.61). The rate of change in muscle strength is presented in Table 2.

Table 2. Change in muscle strength as measured by the physiotherapist, using muscle dynamometry*
Measure of muscle strength6 months12 months24 monthsP for interaction
  • *

    Values are the mean per group. The change from baseline for each subject was calculated as an absolute percent change (compared with the maximum score of 300).

  • From a repeated-measures analysis of variance model. In control patients, recordings of hip flexion and neck flexion were not routinely obtained during the earlier (6 months) period.

Left hip abduction    
 Study group2025260.26
Right hip abduction    
 Study group1217200.69
Left hip extension    
 Study group1722270.03
Right hip extension    
 Study group1723290.07
Left hip flexion    
 Study group152234NA
Right hip flexion    
 Study group172234NA
Left shoulder abduction    
 Study group1717190.74
Right shoulder abduction    
 Study group1620220.61
Left shoulder flexion    
 Study group1318170.47
Right shoulder flexion    
 Study group1620220.61
Neck flexion    
 Study group172633NA
Left grip    
 Study group2023230.20
Right grip    
 Study group2022210.04

All patients in both groups had a rash at the time of diagnosis of DM and initiation of therapy. All patients experienced subjective improvement during the first few months of treatment. At 1 year, however, at least some rash was noted clinically in 14 (47%) of 30 patients in the study group (for whom adequate information was available) and in 9 (43%) of 21 control patients (P = 0.79). At 2 years, 10 (50%) of 20 study patients and 7 (35%) of 20 control patients had some persistent or recurrent rash. Rash was present at 3 years in 3 (33%) of 9 study patients and in 4 (22%) of 18 control patients; the numbers at 4 years were 2 (40%) of 5 study patients and 4 (24%) of 17 controls. The presence of rash was not statistically significantly different between the groups.

During the followup period, calcinosis developed in 5 control patients and 2 study patients. In the 5 control patients, calcinosis was first noticed clinically at 2, 12, 27, 32, and 44 months after initiation of treatment, respectively. In the 2 study patients, calcinosis was first noticed at 5 and 16 months, respectively. The time to development of calcinosis was similar in the 2 groups (for controls versus patients, RR 1.7, 95% CI 0.8–4.4). Over the course of the study, 7 study patients and 7 control patients had disease flares. The time to development of a flare was similar in the 2 groups (RR 1.04, 95% CI 0.61–1.80). The time to development of a flare in both groups is shown as a Kaplan-Meier survival plot in Figure 3. There was no apparent difference between the groups in the response to re-treatment following a flare.

Figure 3.

Survival plot showing time to development of disease flare. The y axis shows the proportion of subjects who were free of disease flare at any point in time. The x axis shows the number of months since starting therapy.

Additional medications.

As shown in Table 3, 15 (48%) of 31 study patients and 6 (27%) of 22 control group patients were also treated with IVIG (P = 0.12). The number of patients receiving any additional immunosuppressive agent was similar in the 2 groups (55% of study patients and 46% of controls; P = 0.50).

Table 3. Additional medications required by study subjects*
  • *

    Values are the number (%). Not all study subjects were followed up for 4 years. NA = not applicable.

Intravenous immunoglobulin15 (48.4)6 (27.3)0.12
Azathioprine1 (3.2)00.40
Cyclosporin A1 (3.2)00.40
Cyclophosphamide1 (3.2)1 (4.6)0.80
Hydroxychloroquine6 (19.4)8 (36.4)0.17
MethotrexateNA6 (27.2)NA
Total no. of subjects receiving any of the above17 (54.8)10 (45.5)0.50

Both of the patients who were treated with cyclophosphamide had ulcerative juvenile DM, characterized by necrotic skin ulcers presumed to be attributable to vasculitis. After receiving 7 monthly infusions of cyclophosphamide, both patients were treated with MTX. Although these 2 patients were treated similarly, they were assigned to study and control groups based on the period during which they were followed up (using the intent-to-treat principle); exclusion of these 2 patients did not change our results.

Eight patients in the control group were treated with hydroxychloroquine, compared with 6 patients in the study group. Six of the control patients were eventually treated with MTX after demonstrating steroid dependence and persistent disease activity.

MTX dosing and side effects.

In the study group, the mean starting dose of MTX was 15 mg (15 mg/m2/week [0.56 mg/kg/week]; minimum 0.19 mg/kg/week, maximum 1.0 mg/kg/week). The maximum dosage during treatment was, on average, 0.64 mg/kg/week (minimum 0.19 mg/kg/week, maximum 1.25 mg/kg/week). The average dosage of MTX at the time of discontinuation of prednisone was 0.53 mg/kg/week. When disease activity was controlled and prednisone could be tapered, MTX was also tapered, with the goal of determining the lowest effective dosage. During additional followup (past the study period), 18 of the 31 patients were able to discontinue MTX while disease was in complete remission. The median length of time receiving MTX was 45.5 months (Figure 4). Seventeen of the study patients received MTX parenterally (subcutaneously) at some point during the course of treatment.

Figure 4.

Survival plot showing time to discontinuation of methotrexate (MTX) for the study group patients. The y axis shows the proportion of subjects who were still receiving MTX at any point in time. The x axis shows the number of months since starting therapy.

Elevations in the levels of liver-related enzymes were observed in 6 study patients and 3 control patients during followup. In 5 of the 6 study patients and in all 3 control patients, these elevations coincided with a flare of disease activity and therefore likely represented muscle inflammation rather than hepatocellular injury. In 1 study patient MTX was discontinued because of persistently elevated levels of transaminase. A liver biopsy was performed in that patient, which revealed nonspecific inflammation. Therapy with MTX was not recommended, however, because the patient's disease was under good control with only corticosteroids. None of the other patients discontinued MTX because of elevations in the levels of liver-related enzymes, and none has developed any clinical signs of liver toxicity.

There were no other significant GI side effects in any of the patients receiving MTX. Infections developed in 3 patients while they were receiving MTX; these infections included 1 case each of shingles, fungal vaginitis, and cellulitis of the metacarpophalangeal joint. Unfortunately, because infections were not as carefully tracked during the time period before MTX was introduced into the clinic, comparisons could not be made. Kawasaki disease was diagnosed in 1 patient during MTX therapy, and the patient was treated accordingly.


Low-dose MTX along with a course of rapidly tapered prednisone, when used as initial therapy for juvenile DM, resulted in good control of disease activity as well as a reduction in the cumulative dose of corticosteroids. This therapeutic approach was associated with fewer corticosteroid side effects, and significant side effects of MTX were not observed.

MTX appears to be safe for use as first-line therapy in juvenile DM. The major concerns regarding MTX include hepatotoxicity, an increased risk of infection, and GI upset (28). In our study, there was no apparent increase in hepatotoxicity or severe GI upset in the group receiving MTX. Overall, the levels of transaminase elevation did not differ between the 2 groups. One patient receiving MTX discontinued the medication because of persistently elevated transaminase levels, but a liver biopsy revealed no significant pathology. Liver biopsies were not warranted in any of the other patients in the study.

In our study, patients were followed up for a relatively short period of time. In the study group, we were able to successfully wean patients off prednisone at a much quicker rate than had been attempted in the control group, and therefore the cumulative prednisone dose and the median time receiving prednisone were significantly less in the study group. Correspondingly, there was less of an increase in BMI, greater height velocity during the early period, and a decreased rate of cataract development in the study group patients over the first 2 years. However, it is unclear whether the benefits of this new approach will continue. Also, some patients with juvenile DM have been reported to experience slowly progressive late morbidity and disability (29). It is possible that our MTX-treated patients will have a different long-term experience than the traditionally treated group. Longer followup is required to identify possible long-term benefits of reduced prednisone exposure, such as changes in bone density and fracture rates.

It is certainly possible that not all of our patients need additional early treatment with MTX; there may be subgroups of patients who would tolerate rapid withdrawal of corticosteroids without any additional treatment. Indeed, a subgroup of 3 of 7 patients from British Columbia who had mild disease were treated with intravenous methylprednisolone infusions on a regular schedule and never needed daily corticosteroids (30). In fact, it is possible that MTX did not add any additional benefit; we may have been able to treat control patients with more rapid corticosteroid tapering without compromising their care. However, it is traditionally believed that shorter courses of corticosteroids result in inadequate disease control and the development of complications (6, 29). Our rationale for using MTX to treat all new patients with juvenile DM was 2-fold: we currently do not have accurate methods for determining which patients with juvenile DM will have a milder disease course, and MTX appears to be very safe when used in low doses to treat rheumatic diseases.

It must be noted that our study patients were treated more often with IVIG than were control patients, although the difference was not statistically significant. At our center we often use IVIG to treat steroid-resistant disease (31). The aforementioned discrepancy is most likely attributable to undertreatment in the control group due to a worldwide shortage of IVIG products during that time period. However, it is certainly possible that IVIG was used more often in the study patients in response to features of disease activity (e.g., continued rash) that resulted from a decrease in the amount of corticosteroids used.

Our results must be considered in light of potential limitations due to our study design. Several of the control patients were eventually treated with MTX. Such treatment represents a form of contamination: by including these children we may have underestimated the potential steroid-sparing effect of MTX. Moreover, we used historical controls, and this method may have introduced bias. There may have been temporal changes in the types of patients presenting to our clinic or changes in other aspects of treatment rather than the MTX protocol. These factors cannot be accounted for unless contemporaneous controls are used. However, our treatments were based on standardized protocols, and data were carefully collected onto standardized forms at the time patients were seen. Also, the same staff at a single center followed up the 2 groups of patients. We believe that we have minimized the potential bias associated with comparisons using historical controls.

Our study should be considered a preliminary step in the assessment of the role of MTX in the initial management of juvenile DM. The results of our study suggest that patients treated with MTX and a course of aggressively tapered prednisone have less exposure to corticosteroids and therefore may experience less associated toxicity. In conclusion, MTX in conjunction with an aggressively tapered course of prednisone may be as effective as traditional long-term corticosteroid therapy while decreasing the cumulative dose of corticosteroids in these patients.