Biologic therapies for refractory juvenile dermatomyositis: five years of experience of the Childhood Arthritis and Rheumatology Research Alliance in North America

Background The prognosis of children with juvenile dermatomyositis (JDM) has improved remarkably since the 1960’s with the use of corticosteroid and immunosuppressive therapy. Yet there remain a minority of children who have refractory disease. Since 2003 the sporadic use of biologics (genetically-engineered proteins that usually are derived from human genes) for inflammatory myositis has been reported. In 2011–2016 we investigated our collective experience of biologics in JDM through the Childhood Arthritis and Rheumatology Research Alliance (CARRA). Methods The JDM biologic study group developed a survey on the CARRA member experience using biologics for Juvenile DM utilizing Delphi consensus methods in 2011–2012. The survey was completed online by the CARRA members interested in JDM in 2012. A second survey was similarly developed that provided more opportunity to describe their experiences with biologics in JDM in detail and was completed by CARRA members in Feb 2013. During three CARRA meetings in 2013–2015, nominal group techniques were used for achieving consensus on the current choices of biologic drugs. A final survey was performed at the 2016 CARRA meeting. Results One hundred and five of a potential 231 pediatric rheumatologists (42%) responded to the first survey in 2012. Thirty-five of 90 had never used a biologic for Juvenile DM at that time. Fifty-five of 91 (denominators vary) had used biologics for JDM in their practice with 32%, 5%, and 4% using rituximab, etanercept, and infliximab, respectively, and 17% having used more than one of the three drugs. Ten percent used a biologic as monotherapy, 19% a biologic in combination with methotrexate (mtx), 52% a biologic in combination with mtx and corticosteroids, 42% a combination of a biologic, mtx, corticosteroids (steroids), and an immunosuppressive drug, and 43% a combination of a biologic, IVIG and mtx. The results of the second survey supported these findings in considerably more detail with multiple combinations of drugs used with biologics and supported the use of rituximab, abatacept, anti-TNFα drugs, and tocilizumab in that order. One hundred percent recommended that CARRA continue studying biologics for JDM. The CARRA meeting survey in 2016 again supported the study and use of these four biologic drug groups. Conclusions Our CARRA JDM biologic work group developed and performed three surveys demonstrating that pediatric rheumatologists in North America have been using multiple biologics for refractory JDM in numerous scenarios from 2011 to 2016. These survey results and our consensus meetings determined our choice of four biologic therapies (rituximab, abatacept, tocilizumab and anti-TNFα drugs) to consider for refractory JDM treatment when indicated and to evaluate for comparative effectiveness and safety in the future. Significance and Innovations This is the first report that provides a substantial clinical experience of a large group of pediatric rheumatologists with biologics for refractory JDM over five years. This experience with biologic therapies for refractory JDM may aid pediatric rheumatologists in the current treatment of these children and form a basis for further clinical research into the comparative effectiveness and safety of biologics for refractory JDM. Electronic supplementary material The online version of this article (doi:10.1186/s12969-017-0174-0) contains supplementary material, which is available to authorized users.

(Continued from previous page) Results: One hundred and five of a potential 231 pediatric rheumatologists (42%) responded to the first survey in 2012. Thirty-five of 90 had never used a biologic for Juvenile DM at that time. Fifty-five of 91 (denominators vary) had used biologics for JDM in their practice with 32%, 5%, and 4% using rituximab, etanercept, and infliximab, respectively, and 17% having used more than one of the three drugs. Ten percent used a biologic as monotherapy, 19% a biologic in combination with methotrexate (mtx), 52% a biologic in combination with mtx and corticosteroids, 42% a combination of a biologic, mtx, corticosteroids (steroids), and an immunosuppressive drug, and 43% a combination of a biologic, IVIG and mtx. The results of the second survey supported these findings in considerably more detail with multiple combinations of drugs used with biologics and supported the use of rituximab, abatacept, anti-TNFα drugs, and tocilizumab in that order. One hundred percent recommended that CARRA continue studying biologics for JDM. The CARRA meeting survey in 2016 again supported the study and use of these four biologic drug groups. Conclusions: Our CARRA JDM biologic work group developed and performed three surveys demonstrating that pediatric rheumatologists in North America have been using multiple biologics for refractory JDM in numerous scenarios from 2011 to 2016. These survey results and our consensus meetings determined our choice of four biologic therapies (rituximab, abatacept, tocilizumab and anti-TNFα drugs) to consider for refractory JDM treatment when indicated and to evaluate for comparative effectiveness and safety in the future.

Significance and Innovations
This is the first report that provides a substantial clinical experience of a large group of pediatric rheumatologists with biologics for refractory JDM over five years. This experience with biologic therapies for refractory JDM may aid pediatric rheumatologists in the current treatment of these children and form a basis for further clinical research into the comparative effectiveness and safety of biologics for refractory JDM.
Due to the above morbidity and the new science, the use of these biologic medications for myositis has begun on an off-label basis since 2000 in numerous countries. The most commonly used biologic in the last decade for inflammatory myositis appears to be rituximab. The Rituximab in Myositis (RIM) study of rituximab effectiveness in adult and pediatric myositis did not meet its primary endpoint for JDM, though the study did show that 83% of the adult and pediatric myositis patients met the definition of improvement on rituximab. Further analysis of the RIM data may allow approval of rituximab in the future [27,28]. Infliximab, an anti-TNF-α medication, has also shown considerable potential for juvenile and adult myositis [29][30][31][32]. In contrast, another anti-TNFα drug, etanercept, does not appear to be as effective for myositis [33][34][35]. The use of a third anti-TNF-α drug, adalimumab, has not yet been reported in myositis treatment but has been helpful for interstitial pneumonitis associated with adult dermatomyositis, anti-synthetase syndrome, and orbital myositis [36][37][38].
Also abatacept, an anti-CTAL-4 monoclonal, has been beneficial in an adult overlap myositis resistant to other treatment [39], in refractory polymyositis [40], and in a recalcitrant JDM child with ulcerations and calcinosis [41]. The anti-IL-6 drug tocilizumab has shown potential for treatment of myositis as it has in other rheumatic diseases [42][43][44]. Other potential biologics for inflammatory myositis are sifalimumab (anti-IFNα), alemtuzumab (anti-CD52), eculizumab (a terminal complement inhibitor) and basiliximab (anti-CD25), but none has not been used much in children as of 2017 [45].
We believed that there was ample support in the medical literature for use and study of biologics off label for refractory JDM. The goal of this article is to report the results of a project performed by our North American Childhood Arthritis and Rheumatology Research Alliance (CARRA) research committee on biologics for refractory JDM.

Carra
CARRA is an organization of pediatric rheumatologists, researchers, and other interested parties in North America that was formed in 2002. Its mission is to prevent, treat, and cure rheumatic diseases of children and adolescents by facilitating and conducting high-quality clinical, translational, and bench research. By 2016 CARRA has grown to have over 400 members in 80 centers in the United States and Canada.

Surveys
Two surveys on the use of biologics for JDM were developed in 2011-13 by consensus discussion of the Juvenile DM biologic committee by e-mail. The first survey of 15 questions that was sent out in February-March 2012 was a general survey on biologics with limited choices for respondents (Additional file 1: Appendix A). It was sent to the general CARRA membership at the time. The results of the first survey are in the results section. A second survey was developed and completed in 2013. The second survey focused on more specifics of the use of biologics with the opportunity for extended comments and on details of the future CARRA studies (inclusion and exclusion criteria, testing, and outcome measures-results not reported here). The second survey results are in Additional file 1: Appendix C. Thirty pediatric rheumatologists who attended JDM Committee meetings at the 2016 CARRA meetings ranked the biologics anonymously using the rank choice technique to check for any change of opinion on the use of biologics for JDM.

Ethics approval and consent to participate
The authors asked for and received Institutional Review Board approval with a waiver from the Nationwide Children's Hospital IRB (2012) and the University of Missouri Medical Center IRB (2013) for survey #1 and survey #2 respectively. No consents were justified.

Results
The first survey can be found in Additional file 1: Appendix A. The results were as follows: The survey was sent to the entire membership of CARRA in 2012 (250 members at the time) and 104 responded (42%). Ninety-seven (93.3%) of the 104 CARRA members replied that they were interested in biologics in JDM and 7/104 replied No (6.7%) and did not continue the survey. Different numbers of the potential 97 respondents answered every question varying from 40 to 91 based on their choice whether any one respondent wanted to answer that one question.
For those who did not use a biologic for JDM, the reasons varied considerably. Two of the 26 (9%) respondents answered that they did not believe biologics worked for JDM. Three of 26 (13%) respondents did not use biologics due to insurance denial. One of 26 (4%) PR did not use a biologic due to parent denial, 15/26 (64%) due to not being sure a biologic would work, and 2/26 (9%) due to cost of therapy.
Twenty-one of 40 (52.5%) reported that the complication of calcinosis was reduced on a biologic while 11/40 (27.5%) reported that muscle atrophy improved. Seven of 40 (17.5%) respondents noted that lipodystrophy was reduced while being treated with a biologic. Eight of 40 respondents (20%) also believed that osteonecrosis complications were reduced while their patients were on a biologic. Seventeen of 40 (42.5%) believed that a biologic helped the severe ulcerative disease of JDM and 10/40   A majority of the CARRA members answering responded that if the opportunity arose, they would use a biologic for JDM (41/47, 87%) while only 3/47 (6%) said that they would not. (Note: this low denominator may indicate that not all respondents wished to express an opinion on this question). Seventy percent of the respondents (63/90) recommended that CARRA study biologics in Juvenile DM, 6% responded no (5/90), 23% (21/90) said they were not sure and 1/90 (1.1%) answered "other". In contrast, 100% of respondents to Survey #2 recommended CARRA study biologics in JDM.

Survey #2
The second survey's purpose was the following: a) Obtain more details on the current and previous use of biologics for JDM by interested CARRA pediatric rheumatologists including the preferred biologics, rationales for their use, combinations of drugs used with the biologic, and to identify which exact biologic might be the primary or secondary drug to use (See Additional file 1: Appendix B); It was also designed to begin the development of the protocol for any future biologic CTP (consensus treatment plandata not reported in this report).
Review of the results of this survey confirms the diverse and widespread use of biologics within the CARRA group. Rationales for use were documented (Additional file 1: Appendix C). There were no recognizable patterns of which biologics were added to which combination of corticosteroids, immunosuppressant's, and IVIG and for what rationale. The results do show support for the use of rituximab, as well as adalimumab, infliximab, abatacept, tocilizumab, and even etanercept.
During the study group sessions of 2014-2015, the study group used these survey results, the current literature, their experiences, and discussion using the nominal group techniques to decide on the ranking of preferences for biologics for refractory JDM. The rankings for treatment and study in these sessions were rituximab, abatacept, infliximab or adalimumab, and tocilizumab in that order.

Survey #3
At the April 2016 CARRA meeting, the 31 physicians attending the JDM work groups ranked the biologics they would use in 2016 for a refractory JDM child unresponsive to corticosteroids, methotrexate, and IVIG utilizing ranked-choice voting (choice 1 high, 5 low). No IRB approval was needed. Rituximab was ranked first (rank mean 1.2), with abatacept (rank mean 2.4) and tocilizumab (rank mean 2.5) ranked second and third, respectively. Infliximab was the fourth choice (rank mean 3.4) and adalimumab the fifth (rank mean 4.3). The preferences for which biologics to use for refractory JDM appeared to remain essentially unchanged from 2012 to 2016 except for tocilizumab moving to a higher ranking.

Discussion
Children with JDM have an excellent prognosis on current treatment in 2016 [2]. Treatment with topical ointments, corticosteroids (prednisone, methylprednisolone), hydroxychloroquine, methotrexate, azathioprine, cyclosporine, and IVIG in different combination treatment regimens has made a tremendous difference. A minority of children with JDM continue to have a difficult course with complications [2]. This JDM group is the target of our efforts with a goal of optimizing our aggressive therapy.
Biologic therapies have revolutionized the treatment of chronic arthritis in children and adults since the late 1990's with markedly improved outcomes and better function. This impressive leap in outcomes cannot be overemphasized [11][12][13][14]. These drugs have targeted proinflammatory cytokines and their receptors such as tumor necrosis factor, interleukin 1, and interleukin six. Other targets include the B cell antigen CD20 and the CTLA Ig molecule. Drugs that target other molecules such as CD-17, CD-23, CD-52 and other molecules are in development [43,45]. Though these drugs are expensive, they appear to be more effective than any other arthritis and rheumatic disease treatment.  . Rituximab has been used the most and is believed to have definite benefit [27,28]. It was our belief that these biologic drugs might benefit children with refractory JDM and these drugs needed to be evaluated. No randomized clinical trials utilized to provide evidencebased guidelines that would improve clinical outcomes appear likely for this rare disease and trials on the use of one biologic do not help to distinguish the comparative effectiveness of different biologics. We began our surveys on biologics in JDM as a possible precursor to comparative effectiveness research approach (CER) used in our organization that might be able to compare different biologic treatments in a multicenter JDM trial.
Our study developed two surveys initially to capture the experience of our CARRA members with biologics in JDM. Both surveys had a respectable sample of pediatric rheumatologists in North America and notably included PR's who care for a large number of children with JDM. The majority of the PR's surveyed had used biologics for JDM as we suspected and the reasons why the minority had never used biologics for JDM were quite predictable. Most North American PR's surveyed did not start treatment of JDM with biologic monotherapy but only used a biologic after steroids, methotrexate, IVIG or another immunosuppressive in some combination have been tried.
It appears important that the survey respondents believed that the biologics significantly reduced complications, particularly calcinosis, muscle atrophy, contractures, lipodystrophy, and osteonecrosis. Side effects were infrequent with patients on rituximab therapy having the most. The most telling result was that 73% of the respondents indicated that children with resistant JDM appeared to benefit from biologic treatment, supporting our belief that biologics were a logical therapeutic step after failure of corticosteroid, IVIG, and immunosuppressive therapy. Also, 87% respondents in the first survey favored using a biologic for JDM if the opportunity would arise. Finally, 70% of respondents in the first survey and 100% in the second survey recommended that CARRA study biologics in JDM.
The second survey results provides for interested pediatric rheumatologists more detailed and rich information on what pediatric rheumatologists were doing with these biologics for JDM as well as their 2013 preferences for the use of biologics for future study. These opinions provided a starting point for the discussions from 2013 to 2015 at our CARRA consensus meetings. It bears repeating that the involved PR's reviewed the surveys, the medical literature, and their experience in making the choice of the four treatment arms studying biologics in JDM using the consensus methods in face-to-face meetings. Without explicit treatment guidelines, the treatments being used are empirical and vary tremendously rheumatologist-to-rheumatologist. This empirical approach, of course, provides no evidence-based consensus to decide which biologic treatment or treatments are optimal. The best solution may be future multicenter comparative effectiveness research to define guidelines for refractory JDM biologic treatments [46][47][48][49][50].
The third survey's purpose was to continue to see if opinions of the CARRA JDM work group members had changed by 2016 on the biologics to use for refractory JDM. Thirty pediatric rheumatologists filled out the survey in April, 2016. A case was described of a child with JDM who was unresponsive to corticosteroids, methotrexate, and IVIG. The rheumatologists ranked their preferences of biologics using rank-choice voting of 1-5 (1 top preference) and chose rituximab with abatacept and tocilizumab nextno major change. Though etanercept received some votes, the consensus opinion of the group was that it is not likely to be effective for refractory JDM.
There are limitations to our survey approach. There was likely a responder bias as the pediatric rheumatologists who answered our surveys were likely the ones who had the most interest in JDM and biologic treatment for JDM. Other PR's who did not respond to the surveys may have had other opinions that were not captured in these surveys. The surveys also had considerable variation in the number of answers of CARRA respondents to any particular question. This variability was likely due to differences in interest in the biologics and wide variation in experience with these drugs. There were no "hard stops" requiring respondents to answer any of the questions. Each CARRA member could choose which question to answer. The surveys also left out a question on the ability to reduce corticosteroid doses by use of the biologics. Although there were a large number of pediatric rheumatologists who responded to the survey and participated in our discussions, there were those in the groups who likely disagreed with the consensus decisions made. Also, these recommendations could not replace the clinical acumen and judgment of each rheumatologist facing the wide clinical variability of any one JDM patient with resistant disease. The process took four years as the CARRA JDM study group worked on the survey and consensus process as well as constructing a consensus treatment plan which delayed this report.