Updates on Efficacy and Safety Janus Kinase Inhibitors in Juvenile Dermatomyositis

Abstract

Introduction:

Juvenile dermatomyositis (JDM) is a rare autoimmune disease most commonly with proximal weakness due to inflammation and characteristic skin rashes. Most patients have a chronic or polycyclic disease course on standard therapy so better treatments are needed. An interferon signature is well-established in key tissues of JDM. Janus kinase inhibitors (jakinibs), which can decrease IFN signaling, are therefore appealing as a targeted therapy.

Areas covered:

Herein is a review of the growing literature on JDM patients on jakinibs, including specifics of their jakinib exposure, summary of efficacy, disease features and characteristics of patients treated, and safety parameters.

Expert opinion:

The vast majority of refractory JDM patients respond to jakinib therapy, though they have varied features, doses, and previous/concurrent medications, and data is largely retrospective. Jakinibs are an exciting and promising treatment in JDM. Evaluation with larger prospective controlled studies is needed to answer remaining questions about jakinibs in JDM regarding dosing, which JDM patients to treat with jakinibs, potential biomarkers to use, and how best to monitor safety risks in JDM.

1. Introduction

Juvenile dermatomyositis (JDM) is a rare autoimmune disease. The most common clinical features include myositis with weakness, characteristic skin rashes, and vasculopathy [1–3]. While some individuals have a monocyclic or monophasic disease course, most have a chronic or polycyclic (recurrence of disease or flare) disease course [4,5]. Myositis-specific autoantibodies (MSA) identify clinical subsets within juvenile dermatomyositis. The most common MSAs are anti-TIF1 (anti-p155/140) which is associated with photosensitivity and chronic disease course, anti-NXP2 (anti-MJ) which is associated with severe muscle disease, dysphagia, and calcinosis, and anti-MDA5 which is associated with interstitial lung disease, arthritis, and ulceration [1,2,6]. Genetic and environmental factors contribute to the immune dysregulation leading to this autoimmune disease, though individual causative factors have not yet been identified [3,5,7]. JDM is known to have inflammation in affected tissues (ex: muscle, skin) which is thought to be related to the features of the disease [8].

Treatment in JDM has the general goal of decreasing inflammation. As there are only two randomized controlled treatment studies with JDM [9,10], most treatment is guided by cohort studies, case series, case reports, and expert consensus such as consensus treatment plans. Specifically, this includes high dose corticosteroids and methotrexate, sometimes with intravenous methylprednisolone (IVMP) and/or intravenous immunoglobulin (IVIg) [3–5,11]. Multiple disease-modifying anti-rheumatic drugs (DMARDs) including cyclosporine, mycophenolate mofetil, azathioprine, tacrolimus, and cyclophosphamide have reported efficacy in JDM [4,5,11,12]. Additionally, some biologic agents have had reports of efficacy in JDM including infliximab, adalimumab, abatacept, rituximab, and tocilizumab [4,5,11,12]. Despite these treatments, most patients still have longstanding disease. Additionally, many patients have multiple notable adverse effects from the broad immunosuppression particularly with high dose corticosteroids, which are a mainstay of treatment with JDM. High dose corticosteroids are often required at higher doses and longer duration for refractory patients [3–5,7]. These factors all indicate a need for more efficacious therapies, ideally more targeted with less side effects.

There have been multiple reports on use of janus kinase inhibitors (jakinibs) in JDM [13–15]. Even standard consensus treatments are not FDA-approved for JDM, and access to janus kinase inhibitors as treatment for JDM is even more limited. Published reports of JDM patients on jakinibs are generally from compassionate use programs or prescribed off-label. While there have been other reviews on JAK inhibitors in myositis including JDM [13–15], the number of JDM patients on JAK inhibitors in literature has continue to increase dramatically, including a recent retrospective study including 101 patients [16]. The growing literature is piecemeal, but a summary of available literature would better reflect overall response and safety for those considering jakinibs in JDM such as clinicians and regulatory agencies. To better summarize available data from these reports, this review will summarize reported efficacy and safety of jakinibs in JDM and highlight key questions remaining.

2. Interferon in Juvenile Dermatomyositis

The inflammatory processes in JDM have been evaluated through multiple studies of proteins and gene expression, finding many different markers of immune dysregulation. However type I (IFN-alpha, IFN-beta) and type II (IFN-gamma) interferon-related markers have been consistently described to be elevated in multiple studies, not only in peripheral blood but also in muscle and skin [8,14,17,18]. The documented correlation of IFN-regulated markers with disease activity measures emphasizes the importance of IFN in JDM pathogenesis [8,14,19–22]. As increased IFN-signaling had been prominently identified in JDM, it has been considered a treatment target that could potentially lead to more effective disease management [3,4,13,14]. As IFN signaling requires janus-kinases (see section 3), a JAK inhibitor is a reasonable targeted treatment option in JDM.

3. Interferon signaling through Janus kinases

Type I and type II IFNs are amongst cell-signaling protein molecules or cytokines that utilize janus kinase (JAK) / Signal Transducers and Activators of Transcription (STAT) (Figure 1) signaling to transduce that cytokine’s signal to the nucleus [23]. Briefly, when these cytokines, including IFNs, bind their receptor, receptor-associated JAKs are activated by phosphorylation. Then, receptor-associated STATs phosphorylate and dimerize, and then translocate to the nucleus and induce transcription of specific genes in response to that particular cytokine [13,14,23]. With IFN, the JAK-STATs induce transcription of IFN-regulated genes. A JAK inhibitor could be effective at decreasing IFN-signaling and downstream expression. There are four JAKS: JAK1, JAK2, JAK3, and TYK2. Type I IFN signals through IFNAR, which utilizes JAK1 and TYK2 while type II IFN signals through IFNGR, which utilizes JAK1 and JAK2 [13,14,23]. JAK inhibitors have varied selectivity per JAK (Figure 1). Any JAK inhibitors of JAK1 or TYK2 would inhibit type I IFN signaling and any JAK inhibitors of JAK1 or JAK2 would inhibit type II IFN signaling.

Figure 1:

Cell-signaling molecules grouped by receptor class and corresponding Janus kinase (JAK) inhibitors.

From the top, there is the receptor class, then specific Janus kinases (JAKs) and associated cell-signaling molecules (cytokines). Then below, JAK inhibitors (jakinibs) are grouped which JAKs are affected, with the less selective first generation pan-jakinibs at the top and more selective jakinibs below. EPO, erythropoietin; IFN, interferon; ILC2, type 2 innate lymphoid cell; CSF, colony stimulating factor; G-CSF: granulocyte-CSF; GH, growth hormone; GM-CSF: granulocyte-macrophage-CSF; OSM, oncostatin M, PRL: prolactin; TPO: thrombopoietin; TSLP: thymic stromal lymphopoietin.

4. Methods

Systematic literature review was done using PubMed including all full publications (excluding conference abstracts) before July 7, 2023. The following search terms (“janus kinase,” “tofacitinib,” “baricitinib,” “ruxolitinib,” “upadacitinib,” “abrocitinib,” “filgotinib,” and “peficitinib”) were combined individually with “dermatomyositis.” We did not specify “juvenile” or “pediatric” in search terms in order to be more inclusive. Juvenile patients from cohorts with mixed adult and juvenile studies were included. Titles and abstracts were reviewed. Papers pertaining only to patients >18 years of age with adult myositis, not related to JAK inhibitor treatment for juvenile dermatomyositis, or not available in English were excluded. All aspects of the review were performed by HK.

5. Overall summary of JAK inhibitor Use in JDM

162 total unique JDM patients from 19 publications on JAK inhibitor or jakinib treatment were identified, with the first publication from 2019 (Table 1) [16,24–41]. Of these, there were eight cases reports with one patient each, three case series with two patients each, six retrospective studies with 9-101 patients, and two open-label prospective studies (three or four patients). Two reports (one case report, one retrospective study) were included in larger published retrospective studies and were excluded from unique patient counts [24,27]. There were no randomized or placebo-controlled studies. 94 (58%) were female from 3-26 years old (all diagnosed < 18 years old). 100 (62%) were on tofacitinib, 30 (19%) were on baricitinib, and 31 (19%) were on ruxolitinib, with one on unspecific jakinib treatment, with varied dosing. The types of patients that were treated, how patients were followed (interval, assessments performed) was variable. Also, monitoring of safety while using the JAK inhibitors or jakinibs and assessment of IFN-signaling markers was not consistent. Thus overall, the quality of data available for use of JAK inhibitors in JDM was not high with 96% retrospective data, 4% open-label prospective studies, and no randomized, blinded, or controlled studies. Nonetheless, use of JAK inhibitors based in JDM is exponentially increasing. Below we will summarize different aspects of jakinib treatment in JDM assessing overall trends from unique JDM patients.

Table 1.

Overview of jakinib publications in JDM

Publication	Study Design	n	# Female	Age (years)*	Disease Duration at Jakinib start (months)*	Primary Country of Publication	Jakinib
Papadopoulou et al (2019)	case report	1	0	11	108	United Kingdom	baricitinib
Sabbagh et al (2019)	case series	2	1	13, 15	18 (NR, n=1)	United States	tofacitinib
Sozeri et al (2020)	case series	2	1	7, 9	42, 52	Turkey	tofacitinib
Kim et al (2020)	open-label prospective study	4	3	5-20	14-110	United States	baricitinib
Yu et al (2020)	open-label prospective study	3	2	10-11	12-72	China	tofacitinib
Heinen et al (2021)	case report	1	0	14	120	Germany	ruxolitinib
Le Voyer et al (2021)	retrospective study	10	7	5-15	0-40	France	baricitinib (3), ruxolitinib (7)
Aeschlimann et al (2018)†	case report	1	1	15	22	France	ruxolitinib
Zhou et al (2021)	case report	1	0	4	24	China	tofacitinib
Min et al (2021)	retrospective study	2	2	19, 26	NR	United States	tofacitinib
Chan Ng et al (2021)	case report	1	0	7	0	Singapore	tofacitinib
Kostik et al (2022)	case series	2	2	6, 10	12, 24	Russia	tofacitinib
Huang et al (2022)	retrospective study	101	58	7.6 +/− 3.7	20 (IQR 12-40)	China	tofacitinib (77), ruxolitinib (23), 1 unknown
Ding et al (2021)‡	retrospective study	25	11	7.2 +/− 4.0	21 (14-37)	China	tofacitinib (7), ruxolitinib (18)
Wang et al (2022)	retrospective study	20	13	3-15	0-42	China	baricitinib
Agud-Dios et al (2022)	case report	1	0	7	24	Spain	tofacitinib
Kaplan et al (2023)	case report	1	1	15	3	Turkey	tofacitinib
Mastrolia et al (2023)	case report	1	1	14	60	Italy	baricitinib
Xue et al (2023)	retrospective study	9	3	8.3 +/− 3.5	38 (mean)	China	tofacitinib
Overall§	8 case reports, 3 case series, 6 retrospective studies, 2 open-label prospective studies	162	94 (58%)	3-26 years	0-110 months	5 China, 3 US, 2 Turkey, 1 each: UK, Germany, France, Singapore Russia, Spain, Italy.	101 tofacitinib, 28 baricitinib, 25 ruxolitinib

^*:Individual ages are listed for n=1 or 2, otherwise the range. If the study only reported summary information, mean +/− standard deviation is noted for age, and for disease duration mean (range) is noted other than for Huang et al (2022), which noted median (interquartile range or IQR) for disease duration at jakinib start

^†:This patient was likely reported in Le Voyer et al (2021).

^‡:These patients were likely reported in Huang et al (2022).

^§:The overall n and jakinib numbers are based on unique patients.

Jakinib = janus kinase inhibitor, NR: not reported, IQR: interquartile range, UK: United Kingdom, US: United States

6. Jakinib effect on Muscle Disease and Function

There are multiple validated muscle measures, including Manual Muscle Testing (MMT) and Childhood Myositis Assessment Scale (CMAS), which are core disease activity measures that are part of the American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) myositis response criteria [5,42–44]. For both, a lower score indicates higher disease activity (more weakness). For this review, active muscle disease was defined by MMT, CMAS, creatine kinase (CK) and/or physician assessment. About one-third of these patients (55/162, 34%), were noted to have muscle involvement at jakinib start (Table 2). One retrospective study noted 10/101 patients had muscle involvement by CMAS (n=6) or CK (n=10) and all improved, though amount and timelines of change were not provided [16]. Overall, 51/53 (96%) had improvement in muscle disease, 35/51 (69%) by MMT or CMAS. Timeline was 1-12 months for 42 patients (not specified for 11 patients) (Table 2). Of the 40 patients with specified timelines, 27 (68%) of them improved in strength at 1-3 months (nine at one month, four at two months, 14 at three months), which is faster than typical response to standard steroid-sparing agents, particularly given that not all were assessed before 3-6 months [12].

Table 2.

JDM muscle disease activity change on jakinibs

Publication	Active muscle disease (n)*	Time to Outcome (months)	n/n improve	How improvement assessed†
Papadopoulou et al (2019)	1/1	6	1/1	MMT, CMAS, CHAQ
Sabbagh et al (2019)	2/2	6	2/2	MMT for 1, CMAS for other; MRI, CHAQ, aldolase for both
Sozeri et al (2020)	2/2	3	2/2	CMAS, CK for both
Kim et al (2020)	2/2	1	2/2	MMT, CMAS, CHAQ/HAQ for both
Yu et al (2020)	3/3	3	3/3	MMT, CMAS, DAS-M, CHAQ for all; CK for 2
Heinen et al (2021)	1/1	3	1/1	CMAS, MRI
Le Voyer et al (2021)	10/10	2-6	8/10	7/9 by MMT, 8/10 CMAS
Aeschlimann et al (2018)‡	1/1	2	1/1	MMT, CMAS
Zhou et al (2021)	1/1	6	1/1	CMAS
Min et al (2021)	0/2	NA	NA	not weak
Chan Ng et al (2021)	1/1	NA	NA	no follow-up on strength
Kostik et al (2022)	1/2	NA	NA	1 weak, no follow-up on strength
Huang et al (2022) §	10/101	19, 12-24§	10/10	6 by CMAS, 4 by exam; CK decreased§
Ding et al (2021)¶	10/25	7, 3-21§	7/9	7/9 by CMAS, no CMAS in 1/10 due to contractures§
Wang et al (2022)	9/20	1-3	9/9	7/9 by MMT, 8/8 CMAS
Agud-Dios et al (2022)	1/1	5	1/1	descriptive, CK
Kaplan et al (2023)	1/1	NR	1/1	cardiac dysfunction stabilized; AST/ALT significantly decreased
Mastrolia et al (2023)	1/1	6	1/1	MRI inflammation score, descriptive
Xue et al (2023)	9/9	2.5-11	9/9	assessment not specified
Overall**	55/162 (34%)	1-24	51/53 (96%)	35/51 (69%) improved with assessment by MMT and/or CMAS. Others: exam (n=4), muscle enzymes (n=2), MRI (n=1). Some supported by description, CHAQ, MRI, DAS-M, and/or muscle enzymes. N=9 assessment unspecified.

^*:Weakness defined as MMT or MMT-8 <78/80 or 135/150 or CMAS <48/52 or subjective. Supportive measures also listed.

^†:Improvement in MMT is an increase of 2+/80 or 3+/150 or CMAS increase of 2+/52 or descriptively.

^‡:This patient was likely reported in Le Voyer et al (2021).

^§:Study only noted timeline as summary (median and range) as listed here. Group muscle disease improvement described in the study. Data not available by individual.

^¶:These patients were likely reported in Huang et al (2022).

^**:This based on unique patients.

AST: aspartate aminotransferase; ALT: alanine aminotransferase; CHAQ: Childhood Health Assessment Questionnaire (/3.0), CK: creatine kinase, CMAS: Childhood Myositis Assessment Scale, DAS-M: Disease Activity Score-Muscle subscore (/9), jakinib: janus kinase inhibitor, MMT: Manual Muscle Testing

Four studies included Childhood Health Assessment Questionnaire (CHAQ)/ Health Assessment Questionnaire (HAQ) [43,44], a patient reported functional health status measure. Eight of 10 patients had a baseline score >0.25/3, with higher scores indicating worse function. Eight of eight decreased by at least 0.25, five at three months and three at six months. Thus, the vast majority of JDM patients treated with jakinibs improve in muscle disease activity and, when assessed, function.

7. Jakinib effect on Skin Disease

Many patients on jakinibs were assessed by validated skin disease activity measures in dermatomyositis including Cutaneous Dermatomyositis Area and Severity Index (CDASI) Activity score, Disease Activity Score – Skin Subscore (DAS-S), and the Cutaneous Assessment Tool – Binary Method Activity Score (CAT-BM) with higher scores indicating higher skin disease activity [5,43,44]. 153/156 (98%) or nearly all of these JDM patients had active skin disease when the jakinib was started; six were missing this information [16,26] (Table 3). For this review, active skin disease [43–45] was defined as CDASI or DAS-S (n=42) , or descriptively (n=14) [28,31,33,40,41]. Two patients did not have any follow-up information on their skin disease [26,29]. Huang et al. [16] used CAT-BM to assess skin disease activity with only group summary statistics were provided without individual change or timeline, but noted 60/96 improved. Including those 60/96, overall 109/152 (72%) had improvement in skin disease activity. Excluding the retrospective large study [16], 49/56 (88%) improved, eight by CDASI, 26 by DAS-S, one by modified skin DAS, and 14 descriptively (Table 3). This change was individually assessed at 1-12 months for 56 patients. Of the 49 that improved with individual timelines, 20/45 (41%) were at 1-2 months (16 by one month, four by two months), and an additional 19 (39%) by three months. Huang et al. noted complete resolution of rash in 16/24 by 8-24 weeks [16]. Many did not have early assessments (before 3-6 months), but as reported with muscle disease activity, some patients had more rapid improvement with rash than other steroid-sparing agents [12]. In summary, there is a clear trend of improvement in skin disease activity in JDM on jakinibs, noted to be relatively quickly for some.

Table 3.

JDM skin disease activity change on jakinibs

Publication	Active skin disease (n)*	Timeline (months)	n/n improve	How improvement assessed†
Papadopoulou et al (2019)	1/1	6	1/1	modified DAS-S, /5
Sabbagh et al (2019)	2/2	6	2/2	CDASI for both
Sozeri et al (2020)	0/2	NA	NA	NA
Kim et al (2020)	4/4	1-3	4/4	CDASI for all, significantly decreased by 1 month, DAS-S for 2
Yu et al (2020)	3/3	3	3/3	DAS-S for all
Heinen et al (2021)	1/1	3	1/1	descriptive (improved)
Le Voyer et al (2021)	10/10	1-3	7/10	DAS-S
Aeschlimann et al (2018)‡	1/1	2	1/1	DAS-S
Zhou et al (2021)	1/1	6	1/1	descriptive (resolved)
Min et al (2021)	2/2	2	2/2	CDASI, patient-reported pruritis improved for both
Chan Ng et al (2021)	NR	NR	NR	NR
Kostik et al (2022)	2/2	NR	2/2	descriptive, 1 improved ulcers
Huang et al (2022)§, ¶	96/96§	19, 12-24§	60/96¶	CAT-BM¶
Ding et al (2021)**,¶	24/25	1	24/24¶	CAT-BM¶, median initial improvement at 1 week**
Wang et al (2022)	20/20	1-3	16/20	DAS-S
Agud-Dios et al (2022)	0/1	NA	NA	NA
Kaplan et al (2023)	1/1	NA	NA	NA – no follow-up noted
Mastrolia et al (2023)	1/1	2	1/1	descriptive (resolved at 9 wks)
Xue et al (2023)	9/9	2-6	9/9	assessment not specified
Overall††	153/156 (98%)	1-24	109/152 (72%)	60 CAT-BM, 26 DAS-S, 8 CDASI, 1 modified DAS-S, 5 descriptive. 9 not specified.

^*:Skin disease activity defined as CDASI >14/100 or modified skin DAS >2/5 or DAS-S of 2+/9.

^†:Improvement in CDASI is a decrease of 5+ or DAS-S or modified skin decrease of DAS 2+.

^‡:This patient was likely reported in Le Voyer et al (2021).

^§:Five individuals not assessed by CAT-BM. Study only noted timeline as summary (median and range) as listed here.

^¶:Group skin disease improvement described in the study. Data not available by individual. Thresholds for active rash and improvement were not specified.

^**:These patients were likely reported in Huang et al (2022). Complete resolution of rash noted in 16/24 by 8-24 weeks.

^††:This based on unique patients.

CAT-BM: Cutaneous Assessment Tool for Myositis – Binary Method (/17), CDASI: Cutaneous Dermatomyositis Area and Severity Index – Activity Score (/100), DAS: Disease Activity Score-Skin (/11), jakinib: janus kinase inhibitor, NA: not applicable, NR: not reported, wks: weeks.

8. Jakinib effect on overall JDM disease activity

Overall JDM disease activity improvement was noted by either ACR/EULAR myositis response criteria (MRC), a score based on change in six individual core set measures (CSM) [5,42,43]. Core set measures include Physician Global Activity (PGA) and Patient Global Activity (PtGA), Extramuscular Disease Activity, Disease Activity Score (DAS), and Physical Summary Score of the Child Health Questionnaire. Overall, 4/4 with muscle and skin disease improved by ACR/EULAR MRC. 23/33 (70%) met varied definitions of complete response or inactive disease at 6 months as specified below.

Kim et al. assessed MRC by both IMACS and PRINTO core set measures [30]. Two patients with muscle and skin disease activity had minimal improvement by week four, and then minimal to major improvement by week eight. For both improved patients, improvement was maintained through week 24 (six months). Sabbagh et al. calculated MRC and noted moderate improvement in two patients also with skin and muscle disease, the timeline was unclear. Thus, those patients with both muscle and skin disease activity (more opportunity to improve) had notable improvement (specifically monitored and sustained in two), as early as week four by MRC [36].

Three studies looked at achievement of inactive disease or similar parameters to assess overall response [43,46,47]. The Paediatric Rheumatology International Trials Organisation (PRINTO) inactive disease criteria includes at least three of the following: CK of 150 IU/I or less, CMAS 48+/52, MMT 78+/80, and PGA by visual analog scale of 0.2/10 or less. Yu et al. noted 3/3 patients met this criteria by six months, after assessment at 3 months [39].

Le Voyer et al. used this criteria at six months, also requiring DAS-S of 0/9 or 1/9 without any skin ulcerations or erythema to define “clinically inactive disease” [32]. Five patients out of ten achieved this, also called clinical response (CR). Two met partial response, with some improvement in CMAS/MMT and DAS-S and decrease of corticosteroid daily dose by at least 50% without any new immunosuppression. One who had a partial response stopped ruxolitinib at 11 months due to a muscle relapse. Another who had a partial response stopped ruxolitinib at 12 months due to insufficient efficacy. Three individuals did not meet partial response and were classified non-responders at six months. Two of these had a severe muscle relapse after a short transient partial improvement while the third had worsening without improvement [32].

Wang et al. utilized similar inactive disease criteria with a higher PGA cutoff of 2/10, and also required DAS-S of 1/9 or less. Three patients of 20 met this criteria (“complete response” or CR) by four weeks, 6/20 had CR by 12 weeks, and 15/20 had CR by 24 weeks [38]. Everyone who achieved CR maintained it through 24 weeks. Three individuals achieved partial response at 24 weeks in Wang et al.’s study [38] with the same definition used as above [32]. Two individuals remained non-responsive by week 24, though one had ILD that improved and the other had calcinosis that remained stable. This indicates specific disease aspects that may have stabilized or improved on baricitinib [38].

Individual core-set measures of interest reflecting overall disease, Physical Global Activity (PGA), and Patient Global Activity (PtGA), were assessed in a minority of studies. These are assessed on a zero to ten visual analog scale, with higher values reflecting higher disease activity [43,44]. Thirty-two individuals from seven studies assessed PGA [24,30,35,36,38–40]. Baseline values ranged from 3-10/10 and decreased to 0-6/10. 30/32 (94%) individuals had a decrease of at least 0.5/10 at 3-12 months. Two individuals remained stable at 5/10 after three months [38]. Twenty-nine individuals from five studies assessed PtGA [24,30,35,38,39]. One individual had a PtGA of 0/10 (PGA of 3/10) [38]. The remainder (28/29) had a baseline PtGA of 3-10/10. All 28 had a decrease of at least 0.5/10 after 2-6 months of jakinib to 0-6/10. Thus, both individual and composite measures of overall clinical response or disease activity generally demonstrate improvement with jakinibs for JDM.

9. Jakinib effect on other specific features of JDM

a. Calcinosis

Calcinosis, or calcium deposits in subcutaneous tissue, fascia, tendons, or muscles, is a comorbidity affecting up to 40% of patients with JDM. It is particularly difficult to treat as it may progress despite improvement other aspects of JDM disease, and can be associated with long-term damage such as contractures, pain, and/or decreased function [3,7,48]. 17/45 (38%) JDM patients on jakinibs were noted to have calcinosis. Overall, nine were noted to have improved, six were stable, and two worsened over 3-12 months (no timeline specified for four patients) (Table 4). As calcinosis is so difficult to treat or manage, lack of progression of calcinosis (stable calcinosis) is generally considered a success. 15/17 (88%) were noted to have stable or improved calcinosis on jakinib. Those with improvement were assessed by x-ray in three [37,38] and clinically in three [25,36] (one later also by x-ray [40]); three didn’t specify how the calcinosis was assessed [32,41]. For those that were stable, two were assessed by x-ray [38], and one by MRI [33], and three were assessed clinically [30,35]. For the two with calcinosis that worsened (new lesion), one was assessed clinically [28], and one did not have a specified assessment [24,32]. Given how difficult calcinosis can be to treat, this points to promise for this treatment in refractory patients, consistent with others who have noted the potential of janus kinase inhibitors in calcinosis [49,50].

Table 4.

JDM-related calcinosis change on jakinibs

Publication	Calcinosis (n)	Timeline (months)	n/n stable or improved	How calcinosis assessed
Papadopoulou et al (2019)	1/1	6, 18	1/1	stable (clinical)
Sabbagh et al (2019)	1/2	12	1/1	improved (n=1) (clinical, detailed body surface area assessment)
Sozeri et al (2020)	2/2	3	2/2	both improved (XR)
Kim et al (2020)	2/4	6	2/2	stable 2/2 (clinical)
Yu et al (2020)	NA	NA	NA	NA
Heinen et al (2021)	1/1	5.5	0/1	new calcinosis (clinical)
Le Voyer et al (2021)	2/10	NR	1/2	1 of 2 improved, 1 new lesion (not specified)
Aeschlimann et al (2018)*	1/1	12	0/1	not improved (not specified)
Zhou et al (2021)	1/1	6, 17	1/1	improved (clinical - 6 mon, XR-17 mon)
Min et al (2021)	NA	NA	NA	NA
Chan Ng et al (2021)	NA	NA	NA	NA
Kostik et al (2022)	NA	NA	NA	NA
Huang et al (2022)†	NA	NA	NA	NA
Ding et al (2021)	NA	NA	NA	NA
Wang et al (2022)	3/20	NR	3/3	1 improved, 2 stable (XR)
Agud-Dios et al (2022)	1/1	~6	1/1	improved (clinical)
Kaplan et al (2023)	NA	NA	NA	NA
Mastrolia et al (2023)	1/1	6	1/1	stable (MRI)
Xue et al (2023)	2/2	8-10	2/2	2/2 resolved (not specified)
Overall‡	17/45 (38%)	3-18	14/17 (82%)	6 clinical, 5 XR, 1 MRI, 1 clinical and later XR, 4 not specified

^*:This patient was likely reported in Le Voyer et al (2021).

^†:These patients were likely reported in Huang et al (2022).

^‡:The overall n and jakinib numbers are based on unique patients.

Jakinib: janus kinase inhibitor, mon: months, NA: not applicable, NR: not reported, PFT: pulmonary function test, XR: x-ray.

b. Interstitial Lung Disease

Another aspect of JDM that can be challenging to treat is interstitial lung disease (ILD), which can be associated with significant morbidity and rarely even mortality [3,4,7,48]. Twenty of the JDM patients on jakinibs are reported to have ILD, though ILD was not systematically assessed so asymptomatic or mildly symptomatic ILD may have been missed. Fifteen of these patients were anti-MDA5 positive (one co-positive for anti-NXP2), two had anti-synthetase autoantibodies, and one had anti-Mi2. One did not have a myositis-specific autoantibody reported. 16/20 (80%) had improvement of ILD (n=14) or stable ILD (n=2) on jakinib therapy (Table 5). Four who had improvement of ILD and one who had stable ILD were assessed by both chest CT and pulmonary function tests (ex: diffusing capacity of the lungs for carbon monoxide (DLCO)) [26,36,39]. Six patients had improvement by chest CT only [31,41]. The timeline to improvement was 2-59 months for these 11 patients. The remaining four patients who improved and one that was stable did not have a specified assessment method or timeline [38].

Table 5.

JDM-related interstitial lung disease (ILD) change on jakinibs

Publication	ILD (n)	Timeline (months)	n/n stable or improved	How ILD assessed
Papadopoulou et al (2019)	NA	NA	NA	NA
Sabbagh et al (2019)	2/2	16, NR	2/2	improved: DLCO, HRCT
Sozeri et al (2020)	NA	NA	NA	NA
Kim et al (2020)	NA	NA	NA	NA
Yu et al (2020)	2/3	6	2/2	improved: DLCO, CT
Heinen et al (2021)	NA	NA	NA	NA
Le Voyer et al (2021)	NA	NA	NA	NA
Aeschlimann et al (2018)*	NA	NA	NA	NA
Zhou et al (2021)	NA	NA	NA	NA
Min et al (2021)	NA	NA	NA	NA
Chan Ng et al (2021)	1/1	2	1/1	improved: DLCO, CXR, CT (RP-ILD)
Kostik et al (2022)	1/2	24	1/1	improved: CT
Huang et al (2022)†	NA	NA	NA	NA
Ding et al (2021)	NA	NA	NA	NA
Wang et al (2022)	5/20	NR	5/5	4 improved, 1 stable: HRCT or clinical
Agud-Dios et al (2022)	NA	NA	NA	NA
Kaplan et al (2023)	NA	NA	NA	NA
Mastrolia et al (2023)	NA	NA	NA	NA
Xue et al (2023)	9/9	4-59	5/9	improved: HRCT (3 with RP-ILD did not improve)
Overall‡	20/37 (54%)	2-59	16/20 (80%)	10 HRCT/ CT, 4 DLCO & HRCT/ CT, 5 HRCT or clinical, 1 DLCO, CXR, & CT

^*:This patient was likely reported in Le Voyer et al (2021).

^†:These patients were likely reported in Huang et al (2022).

^‡:The overall n and jakinib numbers are based on unique patients.

CXR: chest X-ray, CT: computed tomography, DLCO: diffusing capacity of lungs for carbon monoxide assessed by pulmonary function tests, jakinib: janus kinase inhibitor, HRCT: high-resolution computed tomography, ILD: interstitial lung disease, NA: not applicable, NR: not reported, RP-ILD: rapidly-progressive interstitial lung disease.

One rare but potentially fatal complication in juvenile dermatomyositis is rapidly-progressive ILD (RP-ILD). This is more common among those with anti-MDA5 and in East Asian populations [51–53]. Xue et al. reported on three patients with RP-ILD; none of them had improvement by high-resolution chest CT [41]. Chan Ng et al reported on one patient with RP-ILD who had improvement. While that case report focused on potential efficacy of tofacitinib, the patient was also on multiple other concurrent medications [26]. All four of the patients with RP-ILD had anti-MDA5 autoantibodies (one of those that did not improve had both MDA5 and NXP2 autoantibodies). Thus, there is promise for jakinibs for ILD, but broader studies should be performed.

10. Jakinibs for refractory versus treatment-naïve JDM

159/162 (98%) of JDM patients on jakinibs were refractory, defined by insufficient response to previous treatment requiring treatment change, or as described by individual publications. Disease duration was between one month and ten years (not reported in 3 refractory individuals) (Table 1). Huang et al. (n=101) only noted summary data with disease duration of median 20 months, with an inter-quartile range (IQR) of 12-40 months [16]. Xue et al (n=9) noted mean 3.2 years disease duration at jakinib start [41]. 148 cases had at least some individual information on specific previous treatments. All (148/148) were on some form of steroids prior to jakinib therapy and 12/59 (20%) were on hydroxychloroquine. Huang et al. (n=101) did not report specifically on previous hydroxychloroquine. 145/148 (98%) were on steroid-sparing therapies (1 to 10) in addition to steroids with or without hydroxychloroquine prior to jakinib therapy. This included non-biologics (methotrexate (n=46), mycophenolate mofetil (n=21), cyclosporin (n=19), cyclophosphamide (n=18), plasma exchange / immunoadsorption (n=13), tacrolimus (n=5) and one each of azathioprine, leflunomide, sirolimus, thalidomide), as well as biologics (intravenous immunoglobulin (IVIG) (n=101), rituximab (n=17), infliximab (n=5), abatacept (n=4), and one each of adalimumab, anakinra, etanercept biosimilar). Two publications were missing prior medications on 13 JDM patients [16,31].

140 JDM patients had some information on medications concurrent with jakinib therapy. These ranged from just corticosteroids (n=13/140, 9%), just hydroxychloroquine (n=2/57, 4%) or those two together (steroids and hydroxychloroquine, n=2/57, 4%). Huang et al. (n=101) did not specifically report on concurrent hydroxychloroquine. The rest (n=123, 88%) with were on one to five concurrent medications again including non-biologics (methotrexate (n=25), cyclophosphamide (n=9), mycophenolate mofetil (n=7), cyclosporin (n=7), tacrolimus (n=6), plasma exchange / immunoadsorption (n=5) and thalidomide (n=1)) and biologic agents (IVIG (n=64), rituximab (n=9), one each of infliximab, anakinra, and tocilizumab). Five publications did not fully specify individual concurrent medications taken with the jakinib for 24 patients [16,31,35,36,40].

From the refractory patients, most were able to wean steroids with some able to discontinue steroids. 48/52 (92%) of those on steroids with information on steroid weaning were able to wean them, with two noted to discontinue steroids fully. Huang et al. reported that 38/96 (40%) on steroids were able to discontinue them by the last follow-up without reporting whether others could wean without discontinuing [16]. Thus overall, at least 86/148 (58%) were able to wean steroids, of which 40/148 (27%) who were able to discontinue steroids completely. Four patients from two reports were not on daily oral corticosteroids [34,37] with the jakinib. Steroid weaning status was not specified for two patients from two other reports [25,40]. Regarding other immunosuppression, 20/35 (57%) patients on other immunomodulatory medications were able to wean them. In addition to the four excluded as they were only on corticosteroids separate from the jakinib, there were 20 JDM patients who did not specify whether other immunomodulatory therapies were weaned [16,29,31,32,35,37,40,41].

Three treatment naïve patients with more severe disease from three separate studies were given a jakinib in conjunction with other treatment [26,32,38]. Two were anti-MDA5 and one was anti-PL-12. All three had arthritis, two had skin ulcerations, and two had ILD (one of which was RP-ILD). Two had macrophage activation syndrome. Two were given baricitinib and one was given tofacitinib. All three had concurrent corticosteroids and IVIG. Two also received rituximab, and one each was treated with methotrexate, anakinra, cyclosporine, and plasma exchange. Thus, all were on three or four immunomodulatory medications in addition to corticosteroids and the jakinib. The ILD and RP-ILD as well as MAS improved for all who had those features [26,32,38]. JDM disease activity improved in all three, and all were able to wean immunosuppression. However, given that multiple medications were given concurrently in these treatment-naïve patients, it is difficult to assess added or individual benefit from the jakinib.

Thus, many of the patients who were had reports on jakinibs were quite refractory, but the majority were able to wean corticosteroids and/or other agents. This may be due to a more targeted inhibition of IFN-signaling in JDM having more efficacy in managing disease than other immunomodulatory medications. Jakinibs also seem to be beneficial for treatment-naïve severe JDM with given concurrently with standard therapy in a limited number of patients, however assessment of potential greater or more rapid improvement with jakinib added would require further study. Dr. Bader-Meunier is leading a study currently recruiting in Paris, France on baricitinib of new-onset-JDM (MYOCIT) with open-label baricitinib in treatment-naïve patients (n=16) which will hopefully provide additional insight (NCT05524311).

11. JDM on Jakinib: Any differences by MSA Subgroups

There is general interest about which subgroups of JDM may best respond to JDM. One way to assess this is by myositis-specific autoantibody (MSA) group. From reports with individual disease activity data available (n=56), 18 had anti-MDA5, one was anti-MDA5 and anti-NXP2, 15 were anti-NXP2, eight had anti-TIF1, three were anti-synthetase (PL-12 (n=1), PL-7 (n=2)), one was anti-SRP, and one was anti-Mi2. Nine were MSA negative and 5 didn’t report MSA information. Huang et al. noted that from their large study with summary disease activity available, 17 were anti-NXP2, 5 were anti-MDA5, 3 were anti-synthetase (anti-PL7), and 1 was anti-Mi2 with 67 MSA negative by dot blot. Eight did not have MSA information reported.

The two most frequent MSA groups reported on jakinibs are anti-MDA5 and anti-NXP2. Excluding the individual with both anti-MDA5 and anti-NXP2, of those two MSA groups with individual level data and muscle disease activity, 15/15 (100%) of anti-MDA5 [29,32,36,38,39,41] and 11/12 (92%) with anti-NXP2 [24,28,30,32,33,37,38,40] improved versus 51/53 (96%) overall. One of those with anti-MDA5 muscle disease that improved had improvement in cardiac ejection fraction and a decrease in elevated transaminases. Regarding those with active skin disease, 15/16 (94%) anti-MDA5 overlap [26,29,32,36,38,39,41] and 11/14 (79%) anti-NXP2 [24,28,30,32,33,37,38,40] had improvement in skin versus 49/56 (88%) overall.

Overall, there is a clear trend of improvement in JDM muscle and skin disease activity with jakinibs, which remains true within individual MSA groups. The individual MSA subgroups numbers are quite small and assessments were not standardized. It is not possible to assess the disease activity changes by MSA group in Huang et al’s study as only summary activity data was provided. Additionally, there is variability in how MSA testing was performed, and it was not always specified. In at least one study (largest study) [16], MSA testing was done using with dot blot which can have decreased sensitivity for anti-TIF1, one of the most common MSAs in JDM [3,5,54]. Further study is needed with larger cohorts, systemic assessment of patients, and reference standard method MSA testing.

12. Dosing of Jakinibs in JDM

Use of janus kinase inhibitors in juvenile dermatomyositis was part of compassionate use programs (baricitinib prior to any approved indication) and off-label. In general, dosing was guided by other established dosing in pediatric conditions. This included pediatric interferonopathy dosing of baricitinib (4-8mg/day divided) [55], juvenile idiopathic arthritis (JIA) dosing of baricitinib (2 or 4mg/day), JIA dosing of tofacitinib (6.4 to 10mg/day divided), and Graft versus Host Disease (GVHD) dosing of ruxolitinib (12 years old or older, start at 10mg BID, titrate based on hematologic toxicity). Otherwise, three studies [16,27,41] on tofacitinib and ruxolitinib noted dosing was by weight at 2.5mg BID for <25kg and 5mg BID for 25+kg, which is suggested ruxolitinib dosing for children and adolescents for acute GVHD [56].

From JDM reports on baricitinib, five patients from two publications utilized the higher interferonopathy baricitinib dosing starting at 4-8mg/day divided. Three of the patients dose-escalated based on safety and efficacy to 5-12mg/day divided [30,35]. Kim et al. did pharmacokinetic studies that found generally shorter half-life in the lower weight category and longer half-life with lower renal function, similar to what was shown with pediatric interferonopathies [30,55]. Wang et al. noted 11/20 patients were at dosing lower than JIA dosing of baricitinib. They did not do any pharmacokinetic studies or analysis of dose versus response. However, 2/9 (22%) had complete response at JIA equivalent dosing (n=8) or higher (n=1) versus 1/11 (9%) had complete response at lower than JIA equivalent dosing, which may indicate a faster response with higher dose. Additionally, authors noted that both the of the non-responders were at low dosing (both 13 years old on 2mg/day) and that decreased drug dosing may have been influenced by insurance issues for both patients [38].

There were two JDM patients on tofacitinib that were dose escalated to higher dosing. One was a 13-year-old male with JDM who dose escalated from 5mg BID to 10mg BID (20mg/day). This patient developed HSV meningitis which was successfully treated with acyclovir and tofacitinib was not held. Another 26-year-old female with JDM who dose-escalated from 11mg XR daily to 22mg XR daily of tofacitinib. Regarding ruxolitinib, two patients (14-15 years old) started on 15mg BID and one patient started at 20mg BID (11 years old) [28,32]. The 11-year-old was noted to have BK viremia and dose was decreased from 30mg/day to 20mg/day [28]. One patient (15-year-old) who was on 20mg/day of ruxolitinib [24] had pharmacokinetic studies with peak drug levels consistent with those from pediatric interferonopathies and pediatric oncologic studies [57,58].

Dosing and pharmacokinetics of jakinibs should be further investigated in JDM. In some patients, dose escalation was performed based on safety and efficacy. Some JDM patients may benefit from higher dosing, but further study is needed of efficacy versus safety is needed.

13. IFN-signaling markers in JDM on Jakinibs

Given the proposed mechanism of action of jakinibs in JDM, inhibition of pathogenic increased IFN-signaling, several studies assessed different markers of IFN-signaling (IFN-regulated genes, IFN and IFN-related protein, STAT-phosphorylation) in 23 patients (14 with pre/post jakinib assessment) as a proof-of-concept. Three studies noted elevated IFN-regulated gene scores and abnormal STAT phosphorylation (STAT-P) that decrease or reduce to healthy control levels after jakinib treatment in seven patients, who all clinically improved [30,35,36]. Kim et al.’s study also noted CXCL-10/IP-10, an interferon-regulated protein, to decrease in four out of four patients who all clinically improved. Le Voyer et al. reported an elevated IFN-alpha in all ten JDM patients, and a decrease after jakinib in 8/8 of those with follow-up assessment of IFN-alpha. This includes three patients that were non-responders (2 only baseline, 1 with decreased IFN-alpha at follow-up). Thus, baseline IFN-alpha at baseline (high) or trend (decrease with jakinib) did not always correlate with disease activity level or jakinib response [32]. One of the patients from Le Voyer et al. had an earlier case report (11 months of jakinib) was noted to have elevated IFN-regulated gene score and IFN-alpha that remained variably elevated, but STAT phosphorylation that normalized after jakinib therapy [24]. As this patient responded clinically, this indicates that the IFN-regulated gene score and IFN-alpha may remain elevated despite clinical response. Another patient had an elevated IFN-gamma and CXCL10 at baseline which contributed to the decision to include jakinib therapy, without follow-up. This treatment naïve patient responded to jakinib therapy concurrent with other therapies [26].

IFN signaling markers were noted to be elevated in all patients that were tested. Most had markers normalize after JAK inhibitor therapy and most also clinically responded to jakinib therapy. However, there were some exceptions as described. Correlation of these markers with drug exposure from pharmacokinetic studies to assess the relationship to drug dosing and IFN-signaling response in JDM would be helpful. Although IFN-markers generally correlate with disease activity, further study of patients before and after jakinib therapy would clarify how well these markers function reflect disease activity versus jakinib effect. It is possible that these peripheral IFN-signaling markers may not fully reflect the changes occurring in target tissues such as muscle and skin. Thus IFN-markers should be further studied to assess their usefulness in identifying who may respond to jakinibs and monitoring disease activity while on jakinibs.

14. Safety of Jakinibs in Juvenile Dermatomyositis

a. Highlights of Safety of Jakinibs in Other Conditions

In rheumatology, janus kinase inhibitors was first approved for moderate to severe adult rheumatoid arthritis (RA) (tofacitinib, filgotinib, baricitinib, upadacitinib and peficitinib). Ruxolitinib was first approved for myelofibrosis and has other indications including graft versus host disease (Table 6). As the largest amount of information on safety of jakinibs is for adults, we will review that for context as providers and parents commonly ask about these risks. Notably, a black box warning was placed on adult RA jakinibs including tofacitinib and baricitinib based on the phase 3b/4 randomized safety endpoint Oral Surveillance study (NCT02092467) of adults 50+ years old with RA and at least one cardiovascular risk factor for tofacitinib versus tumor necrosis factor inhibitor (TNFi). Briefly, the study found an increased risk of infection, all-cause death, malignancy, major cardiovascular events, and thrombosis with tofacitinib versus TNFi [59,60]. However, we know this population is already at higher risk for these events with underlying inflammation and age-related risk, and there was no control group which complicates interpretation. Further analysis is needed to assess inherent differences between those given tofacitinib versus a TNFi. Based on the same study information, given the absence of evidence for greater risk of tofacitinib versus TNFi in patients without risk factors, EULAR recommendations include that a jakinib can be considered for RA after failure of a conventional disease-modifying anti-rheumatic drug (DMARD) treatment after evaluating relevant risk factors. Additionally, it is unknown whether there would be similar risks noted for a similar outcome-based randomized controlled study for DMARDs [61,62]. Similar risks have been noted for tofacitinib, baricitinib, and ruxolitinib in general versus placebo regarding infection, with increased treatment-emergent infections with jakinibs and higher incidence of herpes zoster [62–64]. Also, changes in cell counts, transaminitis, creatine kinase (CK) elevation, and elevated cholesterol, LDL, and/or triglycerides have been observed [65–67].

Table 6.

Approved jakinibs and indications

Drug	JAK Specificity	Approved Indication
Abrocitinib	JAK1	AD
Baricitinib	JAK1, JAK2	RA, JIA (Europe), COVID-19, AA (USA), AD (Europe, Japan)
Delgocitinib	Pan-JAK	AD (topical only, Japan)
Deucravacitinib	TYK2	Psoriasis
Fedratinib	JAK2	MF, PV, ET
Filgotinib	JAK1	RA (Europe, Japan), UC (Japan)
Peficitinib	Pan-JAK	RA (Japan)
Ritlecitinib	JAK3	AA
Ruxolitinib	JAK1/JAK2>TYK2>JAK3	GVHD, MF, PV, ET, AD (topical only), vitiligo (topical only)
Tofacitinib	JAK3/JAK1>JAK2, TYK2	RA, UC; PsA, AS, JIA (USA, Europe)
Upadacitinib	JAK1>JAK2	RA; PsA, AD (USA, Japan); AS, UC (USA)

Drugs listed alphabetically.

AA: alopecia areata, AD: atopic dermatitis, AS; ankylosing spondylitis, ET: essential or post-essential thrombocytopenia, GVHD: Graft vs. Host Disease, jakinib: janus kinase inhibitor, JIA: juvenile idiopathic arthritis, MF: myelofibrosis, PsA; psoriatic arthritis, PV: polycythema vera, RA: rheumatoid arthritis, UC: ulcerative colitis

With Mendelian interferonopathies which is predominantly pediatric patients, treatment-emergent adverse events with baricitinib were assessed for 18 patients. Of note, two patients developed herpes zoster. BK-virus related changes were noted on baricitinib, with increases in BK viremia (n=9) noted. One patient developed both BK viremia and azotemia. This study also reported eight individuals had transaminitis and five individuals had elevation in creatine kinase. Eight of 18 patients had weight gain. Neutropenia (n=3), lymphopenia (n=2), thrombocytopenia (n=2), and thrombocytosis (n=4) were also noted [68]. JC virus-related pathology has also been reported with ruxolitinib for other indications [69], and more recently for an adult with systemic lupus erythematosus on tofacitinib [70].

Access to jakinibs for JDM is limited as there are limited pediatric indications in general. Tofacitinib and baricitinib (Europe) have been approved for moderate to severe juvenile idiopathic arthritis (JIA) and ruxolitinib has been approved for adolescent (> 12 years old) refractory graft versus host disease (Table 6). From pediatric studies for those indications, similar risks were observed regarding infections including herpes zoster, herpes simplex, and changes in cell counts and transaminitis [71–73]. From the tofacitinib and baricitinib studies for JIA, there were no major cardiovascular events, malignancy, or death. In the tofacitinib study, there were no thromboembolic events [71]. There was one individual (1%) who was in the baricitinib group who had a pulmonary embolism in the double-blind withdrawal period. This individual had individual risk factors of baseline thrombocytosis, overweight, recent immobilization, high disease activity, suspected pneumonia, but the thrombosis was assessed to be study-drug related [72]. Thus, jakinibs generally are overall well-tolerated in these other pediatric conditions with similar infection and laboratory safety profile. The long-term risks for malignancy, thrombosis, cardiovascular events, and death raised in adults require further study, but are likely lower in pediatric populations who lack many of risk factors studied in the adult RA population.

b. Evidence on Safety of Jakinibs in JDM

Of the 19 publications on jakinibs in JDM, only two were prospective of three and four patients [30,39] and only one assessed all treatment-emergent adverse events systematically in four patients over 24 weeks [30]. For the remaining retrospective studies, there is increased bias in reporting or assessment of safety events, so overall assessment of incidence and frequency of events cannot be done. Overall total time on jakinib per patient is between 3-35 months. Four of these publications did not have any comment on adverse events, infections, or safety [29,31,37,41]. One publication included “no adverse events” without any further specifics [25].

In general, infections were most commonly described. Upper respiratory infections were noted to be the most common infection when reported [16,30,38]. One patient on tofacitinib 10mg BID tofacitinib developed herpes simplex virus (HSV) meningitis which resolved with acyclovir as mentioned above [36]. From Le Voyer et al.’s report, 3/10 had herpes simplex virus infections, one of which was hospitalized, though all continued jakinib therapy [32]. Huang et al.’s retrospective study noted 2 infections with varicella zoster virus (VZV), 5 with Epstein-Barr virus (EBV), and 2 with cytomegalovirus (CMV), though it is unclear if these were symptomatic or based solely on regular laboratory screening [16]. One patient from Wang et al.’s study was noted to have HSV, requiring baricitinib to be held and hospitalization due to this infection [38]. Le Voyer et al. also noted 2 with skin ulcerations with staph aureus abscesses and one psoas abscess [32]. Wang et al. noted two fungal infections [38]. Four other studies indicated no hospitalizations or serious or significant infections [26,33,34,40].

A subset of studies also assessed for BK and/or JC virus. Two studies noted an increase in BK virus. One case report with ruxolitinib noted asymptomatic BK viremia and subsequently decreased ruxolitinib dose [28]. Kim et al. systematically assessed BK viremia and viruria and noted one of four patients to have an increase in BK viremia, and another patient had new BK viruria both not requiring any change in baricitinib [30]. One other case report noted “no BK or JC viremia” [35].

Regarding laboratory changes, three studies noted that there were no cytopenias or hematologic side effects [24,35,40]. Huang et al. noted a significant decrease in white blood cell count with an increase in leukopenia (n=14/95) with tofacitinib or ruxolitinib [16]. Kim et al. found 1/4 with anemia, neutropenia, and lymphopenia, and 2/4 with thrombocytosis with baricitinib. From the four patients, there was a significant decrease in hemoglobin and increase in platelets [30]. One study noted that lipids and creatinine were within normal limit for their case report with tofacitinib [40]. Wang et al. noted 1/20 with elevation of creatinine with tofacitinib or ruxolitinib [38]. Kim et al. noted 1/4 each with elevated triglycerides and high cholesterol. Two of the four patients had weight gain [30].

Regarding muscle enzymes, one retrospective study (n=20) noted one patient with elevation of ALT on baricitinib [38], while another retrospective study (n=25) noted elevation of LDH for two (tofacitinib or ruxolitinib) [27]. One case reported noted liver enzymes were within normal limit for their case report [40]. Regarding muscle enzymes, two were noted to have increased creatine kinase, two with increased LDH, and one with increased aldolase with baricitinib [30]. The inconsistent increases documented in muscle enzymes (transaminitis, LDH, CK, aldolase) are notable as these are typically monitored for muscle inflammation in myositis. However, elevations were not associated with increase in muscle inflammation or weakness in any of the studies. Transaminitis and CK elevation have been previously reported with in other populations on jakinib, while LDH and aldolase are not typically monitored in the other populations on jakinibs. This increases the likelihood that these elevations may be adverse effects of jakinibs. Muscle strength and inflammation should be closely monitored independent of muscle enzymes in juvenile myositis patients on jakinibs.

Finally, there were no thromboembolic events, cardiovascular events, or malignancies reported in any of the JDM on jakinib reports. One retrospective study of 101 patients noted one death due to “ineffective treatment” without other details. There were no other deaths reported. Only one study had systematic prospective monitoring (n=4) and did not include any such events [30]. Limited (four) other studies commented on any of these events explicitly [16,27,34,38] while the rest did not. Overall, the safety profile is similar to what has previously been described in other juvenile cohorts.

15. Conclusion

There is growing literature with more than 150 juvenile dermatomyositis patients reported in the literature on jakinib therapy (ruxolitinib, baricitinib, tofacitinib), though only seven patients from two studies were prospectively studied on open-label drug and none of the studies were randomized, blinded, or controlled. There is a clear trend of clinical improvement in muscle (51/53, 96%) and skin disease activity (109/152, 72%) from primarily refractory (159/162, 98%) patients, primarily by validated outcome measures including patient-reported assessments. The majority of those noted to have calcinosis and ILD were noted to have improvement or stabilization of those features, though this is likely biased by preferential reporting of those who improved. Having jakinibs on-board has allowed for steroid-weaning in most and weaning other medications in some. Varied dosing was used (ex: lower dosing similar to JIA vs. higher dosing similar to interferonopathies) without assessment for optimal dosing in JDM. In the limited patients with IFN-signaling markers assessed, all were increased/abnormal at baseline (n=23) and most decreased/normalized with jakinib treatment (13/14), though this did not fully correlate with clinical response to jakinib therapy.

Safety assessment is limited, but some were noted to have increases in muscle enzymes which did not correspond to increased weakness. In general, the safety profile seems similar to that seen for jakinibs for other pediatric conditions with increased risk for some infections such as herpes simplex and herpes zoster and changes in cell counts.

As the vast majority of these studies were retrospective reports (96%) and none were controlled studies, better quality studies (controlled, larger) are needed to better characterize optimal dosing in JDM, determine which patients may best respond to jakinibs (ex: biomarkers, MSA groups), when during disease course to use jakinibs, and what safety parameters to monitor and how best to counsel patients and families regarding risks and benefits of jakinibs in JDM.

16. Expert opinion

Though standardized treatment has been established for JDM, most patients still have a chronic or polycyclic disease course. This indicates a need for more effective therapy. Additionally, high-dose corticosteroids are a mainstay of therapy and carry a high burden of adverse effects. As JDM has a large body of literature noting an increased IFN signature in key tissues (blood, muscle, and skin) for this disease, this represents a promising treatment target. There are many ways to target IFN, but one that could be utilized off-label currently for JDM is oral janus kinase inhibitors. As this represents a more targeted treatment option in JDM with the goal of inhibiting type I and type II IFN signaling, it has the potential of increased efficacy and decreased side effects for patients.

The growing literature of jakinib use in JDM likely represents enthusiasm of healthcare providers regarding early reports of clinical response in refractory patients. The first five publications in 2019-2020 (13 patients) were all individuals with impressive refractory disease that improved on jakinibs. Though many patients respond to standard therapy in general, reported patients had been on up to ten steroid-sparing agents prior highlight a subpopulation of JDM that needs more efficacious therapy. The unique response to jakinibs for the vast majority of patients indicates the distinct mechanism of action with jakinibs may better target pathology in JDM. This is supported by the decrease in IFN-signaling observed. These reports support jakinibs as an exciting and promising treatment to consider in JDM, though there are caveats from the data quality of primarily retrospective studies. Additionally, other IFN-targeting measures should be considered to expand treatment options in JDM. For example, anti-IFN-beta (PF-06823859) [74] infusions are being studied in refractory adult dermatomyositis and polymyositis as a randomized clinical trial (NCT05895786).

There are many questions remaining regarding jakinib treatment in JDM. The current literature is largely retrospective and biased towards those that respond. As larger studies started being reported in 2021, more non-responders were identified [32,38]. Further study should be done to identify any predictors of response versus non-response. Specifically, it would be helpful to assess if certain disease features, myositis-specific autoantibody groups, age-groups, or points in disease duration are best for jakinib use in JDM. This would require larger prospective studies with reference method myositis-specific autoantibody testing, systematic follow-up, and biosample collection.

Finally, safety is a key question needing more follow-up. Most would agree the safety profile of jakinibs is preferred over that of daily oral corticosteroids. However, this is a newer medication overall and risks raised in adults raises concerns for pediatric patients and families. Monitoring and screening for infection is typical of the immunosuppressive medications commonly used in JDM. Based on data so far, we recommend review of patient history and vaccination as appropriate for herpes zoster and herpes simplex prior to starting a jakinib in JDM. Regular laboratory monitoring of blood counts, liver function tests, renal function, and BK virus in blood and urine should be done. There is limited data, but growth should be monitored as growth hormone utilizes JAK/STAT signaling. Muscle enzymes (transaminases, CK, LDH, aldolase), more specifically assessed in myositis patients such as JDM, have been variably elevated so this should be further monitored with assessment of muscle disease activity by other means. Long term follow-up is needed to better quantify risks including long-term risks such as malignancy, cardiovascular events, and thrombosis.

In summary, jakinibs represent an important addition to the treatment armamentarium in JDM although current literature has significant limitations. Specifically, though many JDM patients on jakinibs are in literature, 96% are retrospective reports and none of the studies are controlled, randomized, or blinded. By inhibiting a treatment target (IFN-signaling) unique from most standard medications, impressive clinical response was noted in primarily retrospective reports of refractory patients. Many aspects of JDM treatment with jakinibs require further study including what may help predict response, potential subgroups including MSA groups that may be more responsive, and risk assessment. Larger prospective studies, ideally controlled, with systematic assessment and follow-up and biosample collection for research evaluation will help us generate high quality data in order to better personalize management of JDM patients with jakinibs.