In this single-center 14-year cohort study, we demonstrate that JDM patients treated with low-dose oral glucocorticoids (with IVMP only if dictated by disease severity) and early use of steroid-sparing treatments have comparable results to published outcomes in other treatment protocols advocating higher doses and longer duration of steroids, including a well-designed, similar sized, single-center cohort study that used a more conventional glucocorticoid regimen (16). Our treatment protocol included intravenous steroid use at diagnosis in only 38% of patients, contrasting with 84% in the comparator single-center cohort study (16). Furthermore, our treatment protocol employed lower starting doses of oral steroids (median dose of 0.85 mg/kg/day) that were tapered over time as the patient responded, in contrast to a more conventional higher dosing regimen of 2 mg/kg/day given until a complete clinic response was achieved (16). We utilized a higher frequency of IVIG (48% versus 20%) and rituximab (16% versus zero); however, this observation must take into account that the study periods do not completely overlap and reports of rituximab use in JDM (25, 26) had not yet been published during the study period of the comparator study (16). Our use of cyclophosphamide in 10% of patients compared to 4% in the comparator study (16) suggests that both studies saw approximately similar disease severity, as cyclophosphamide is usually reserved for the most severely affected JDM patients. When comparing clinical responses to the different treatment regimens, our patients achieved myositis and cutaneous disease control with medians of 7.1 and 16.7 months compared to myositis and cutaneous disease normalization medians of 13 and 19 months, respectively, in the comparator study, although the definitions used were slightly different. Sustained medication-free disease control in both cohorts was comparable, with 21 patients (68%) achieving medication-free disease control in our study, compared to 28 patients (57%) in the comparator study. Overall, it appears that outcome measures were comparable between the two studies.
Calcinosis is considered a marker of disease damage rather than an active disease feature itself (27, 28), although that understanding may be changing, as a recent survey of pediatric rheumatologists reported 73% of respondents considered the new development of calcinosis in a patient with absent muscle or skin disease as “active JDM disease” (29). In our cohort, calcinosis developed in 19% of patients (6 out of 31 patients), but persisted in only one patient (3%). The calcinosis afflicting five of the six patients resolved with standard treatment prescribed for their JDM, without sequelae by the end of the of the study period. In the comparator study (16), six patients (12%) developed calcinosis, which was persistent in two patients (4%). The incidence and prevalence of calcinosis in our cohort is comparable to the published literature (30), and if calcinosis is understood to be associated with prolonged or inadequately treated disease (31), then these findings also reflect the effectiveness of our treatment approach using lower doses and duration of glucocorticoids.
As described by Stringer et. al., (17), the treatment strategies employed by pediatric rheumatologists in treating JDM are largely anchored around the use of high-dose oral glucocorticoids. This is echoed in the published treatment guidelines of CARRA (20) and SHARE (22) which both advocate for the use of IVMP and oral prednisone starting at 2 mg/kg/day tapered over 12 months. At our center, patients were started at a lower dose of steroids (median 0.85 mg/kg/day), followed by a taper based on the clinical and laboratory response of the individual patient completing in a median 9.1 months (range 4.7–17.4). High-dose IV methylprednisolone was reserved for moderate to severe JDM cases, comprising only 38% of patients. Our alternative approach of lower steroid doses combined with universally early initiation of steroid-sparing agents led to lower cumulative glucocorticoid exposure and less associated adverse effects while retaining comparative favorable outcomes. A previous retrospective study of IVMP therapy compared to oral prednisone (1–2 mg/kg/day) in JDM also found little difference in efficacy after controlling for disease severity (32).
We did not identify a significant difference in glucocorticoid usage between JDM patients with monocyclic and chronic continuous courses. While high-dose intravenous glucocorticoids remain the first-line option for patients with moderate to severe initial presentation, following this initial period, our data suggests that a taper beginning at a lower steroid dose can be used. Furthermore, our retrospective cohort suggests that a sustainable result can be achieved without the need for frequent IVMP pulses following the induction IVMP pulse. In our study, steroid-sparing agents, particularly methotrexate, IVIG and hydroxychloroquine, were used universally early in the disease course and continued after glucocorticoid discontinuation. Monthly IVIG was used with very good results, despite the disadvantages of high cost and the need for an infusion facility. No major adverse events attributable to IVIG were seen in our cohort, possibly influenced by the intentional avoidance of central line placement.
Cutaneous disease is often more resistant than myositis to initial treatment with glucocorticoids and immunomodulators, often persists for longer periods of time, and in some cases is refractory to multiple therapies. Hydroxychloroquine has been reported as an effective agent for refractory cutaneous disease (33, 34) and has been included by CARRA in consensus treatment plans of skin-predominant JDM (35). In our study population, hydroxychloroquine was utilized in 13 patients with persistent skin involvement. Of the 15 cases in which IVIG was used, six cases were to treat refractory cutaneous disease, and was effective in these patients. Four of the five patients in this cohort who received rituximab have been previously reported, and rituximab was found to be beneficial in three of four cases (25). Mycophenolate mofetil has been reported to be a useful steroid-sparing agent in patients with JDM (36), and in this study cohort was used in 2 patients with beneficial results.
Limitations of the study include the retrospective design and the relatively modest number of subjects due to our applied exclusion criteria. Due to the retrospective nature of this study, we were not able to apply PRINTO’s criteria for inactive disease, which includes the childhood myositis assessment scale, manual muscle testing and/or physician global assessments. In reference to the comparator study, some definitions in the two studies were not identical making some direct comparisons difficult. The lack of universal testing of myositis antibodies did not allow sufficient numbers to establish associations between specific antibodies and disease features or outcomes.