Systemic juvenile idiopathic arthritis: Treatment

INTRODUCTION — Systemic juvenile idiopathic arthritis (sJIA, formerly called Still's disease or systemic-onset juvenile rheumatoid arthritis) is a subset of JIA. Adult-onset Still’s disease (AOSD) is probably the same or similar disease when it begins in patients ≥16 years of age. sJIA is classified as a subset of JIA, but the pathophysiology is most consistent with an autoinflammatory disorder. (See "Classification of juvenile idiopathic arthritis", section on 'Systemic arthritis'.)

The treatment options for sJIA are discussed in this review. The clinical manifestations, diagnosis, complications, and prognosis of sJIA are discussed separately, as are other subtypes of JIA. (See "Systemic juvenile idiopathic arthritis: Clinical manifestations and diagnosis" and "Systemic juvenile idiopathic arthritis: Course, prognosis, and complications" and "Oligoarticular juvenile idiopathic arthritis" and "Polyarticular juvenile idiopathic arthritis: Clinical manifestations, diagnosis, and complications" and "Classification of juvenile idiopathic arthritis".)

OVERVIEW — Nonsteroidal antiinflammatory drugs (NSAIDs) alone are effective for some children with mild sJIA [1]. The traditional treatment for patients who failed NSAID therapy was glucocorticoids along with the disease-modifying antirheumatic drug (DMARD), methotrexate. However, glucocorticoids are a poor long-term therapeutic option because of associated toxicity, and methotrexate often does not change the aggressive and often permanently disabling consequences of unremitting disease. Biologic DMARDs, such as interleukin (IL) 1 or IL-6 inhibitors, were initially reserved for patients refractory to conventional therapy (NSAIDs followed by the addition of glucocorticoids with or without methotrexate). However, they are increasingly the agent of choice after failure of NSAID therapy in patients with a confirmed diagnosis of sJIA since they are highly effective.

Treatment guidelines, based upon presence or absence of active systemic features, clinician global assessment, active joint count, and presence or absence of features concerning for macrophage activation syndrome (MAS), are outlined by the American College of Rheumatology (ACR) [2]. These guidelines emphasize the earlier use of biologics in children with sJIA, although specific information on appropriate dose is lacking. Another set of standardized treatment plans was developed through a consensus process by the Childhood Arthritis and Rheumatology Research Alliance (CARRA) based upon the most commonly used treatment approaches for sJIA [3].

OUR APPROACH — Our general approach, which is similar to the American College of Rheumatology (ACR) guidelines and Childhood Arthritis and Rheumatology Research Alliance (CARRA) consensus treatment plans in terms of treatment choices, is reviewed here, and the specific agents are discussed in greater detail below, including dosing and withdrawal of therapy.

Initial therapy — The choice of initial therapy depends upon the severity of the presentation, including whether the patient shows signs of macrophage activation syndrome (MAS). (See "Systemic juvenile idiopathic arthritis: Clinical manifestations and diagnosis", section on 'Clinical manifestations' and "Clinical features and diagnosis of hemophagocytic lymphohistiocytosis", section on 'Clinical features'.)

Mild-to-moderate acute disease — We suggest nonsteroidal antiinflammatory drug (NSAID) monotherapy as the initial treatment in children with possible sJIA who have a mild-to-moderate presentation, meaning nondisabling symptoms (fever, rash, mild-moderate arthritis) without evidence of MAS. Any NSAID other than aspirin is suitable and is given at standard antiinflammatory doses. NSAIDs are especially useful for treatment of disease in the early stages, when the diagnosis may not be clear and the work-up is still in progress. (See 'Nonsteroidal antiinflammatory drugs' below and "Systemic juvenile idiopathic arthritis: Clinical manifestations and diagnosis", section on 'Clinical manifestations' and "Systemic juvenile idiopathic arthritis: Course, prognosis, and complications", section on 'Macrophage activation syndrome'.)

In general, a trial of NSAIDs alone should last no more than a week or two since it is usually quickly apparent whether an additional agent is needed within this timeframe. Additional therapy is indicated in children who develop or continue to have significant symptoms despite use of NSAIDs (such as continued high fevers, rash, arthritis, serositis). The approach for patients who have not responded sufficiently to a trial of an NSAID alone during the acute phase of the illness is similar to that for patients who present with moderate-to-severe disease. (See 'Moderate-to-severe acute disease' below.)

Moderate-to-severe acute disease — For patients whose initial symptoms include serious systemic manifestations such as serositis, possible early MAS, moderate-to-severe polyarthritis (with or without a trial of NSAIDs), or high fevers with a poor response to NSAIDs, we suggest starting one of the biologic agents that inhibit interleukin (IL) 1 or IL-6, such as anakinra, canakinumab, or tocilizumab. NSAIDs can be continued in conjunction with other agents if needed for pain control. The choice of biologic agents should be a decision that is shared with the patient/family and clinician and depends upon multiple factors such as differing routes of administration and frequency of dosing. As an example, anakinra is given as a daily injection, canakinumab as a monthly injection, and tocilizumab as an intravenous infusion every two weeks. Some pediatric rheumatologists prefer starting with a short-acting biologic such as anakinra, especially at the onset of disease, because it can be easily withdrawn if there is a question about the diagnosis or a complication such as infection, but it is not yet approved for use in this disease in the US, making it difficult to obtain at times. (See 'Interleukin 1 inhibitors' below and 'Interleukin 6 inhibitors' below.)

Some pediatric rheumatologists prefer to start with glucocorticoids in this situation and use methotrexate as a steroid-sparing agent since glucocorticoids provide rapid onset of action and almost uniform efficacy for even severely ill children. However, the biologic agents that block the effects of IL-1 and IL-6 are often more effective than nonbiologic disease-modifying antirheumatic drugs (DMARDs; eg, methotrexate) and have more favorable side-effect profiles than long-term use of glucocorticoids. In addition, glucocorticoids should not be used unless the possibility of malignancy, which may cause a similar presentation, has been excluded. (See 'Glucocorticoids' below and 'Methotrexate' below and 'Biologic disease-modifying antirheumatic drug (DMARD) therapy' below.)

Patients presenting with relatively severe symptoms or who have suspected early MAS should be treated with glucocorticoids in addition to an anti-IL-1 or IL-6 agent. Concomitant use of a biologic agent may allow relatively rapid tapering of glucocorticoids. (See 'Refractory acute disease' below and "Systemic juvenile idiopathic arthritis: Course, prognosis, and complications", section on 'Macrophage activation syndrome'.)

Refractory acute disease — For patients started on a biologic DMARD as part of the initial therapy, we suggest adding a glucocorticoid within a week if there is continued polyarthritis, fever, and rash. A glucocorticoid should be added sooner if there was evidence of MAS or severe serositis. Both drugs are then typically continued until disease control is established. The clinician may then decide to gradually withdraw the glucocorticoid first because of the inevitable toxicity associated with its chronic use. Withdrawal of the biologic agent may be attempted as well if the patient continues to exhibit no sign of disease since there are infrequent patients who have monophasic disease, which will remit completely after a period of months [4-6]. Evidence suggests that there is a window of opportunity, and the early use of anti-IL-1 agents may change the course of disease [7,8]. (See 'Glucocorticoids' below and "Glucocorticoid withdrawal" and 'Biologic disease-modifying antirheumatic drug (DMARD) therapy' below.)

Biologic DMARDs are the preferred choice for patients with refractory acute disease who were treated with an NSAID plus a glucocorticoid with or without methotrexate. Anti-IL-1 and anti-IL-6 agents have shown good efficacy and allowed tapering of glucocorticoids in randomized trials [9,10]. (See 'Mild-to-moderate acute disease' above and 'Interleukin 1 inhibitors' below and 'Interleukin 6 inhibitors' below.)

Recurrent or polyphasic disease — Some patients have a polyphasic disease course, with the disease becoming completely quiescent for months or years without treatment, then recurring episodically [5,6]. The approach to treating these recurrent episodes is the same as for initial therapy.

Persistent chronic disease — Patients with persistent chronic disease that recurs after drug tapering or who do not completely respond to a combination of a biologic agent and glucocorticoids plus methotrexate or other DMARDs may have clinical differences that can help decide treatment. As an example, some have primarily systemic manifestations, others primarily arthritis (which can be progressive and destructive), and still others have both. Inhibitors of IL-1 and IL-6 appear to be the most effective of the biologic agents for primarily systemic disease (eg, fever, rash, and serositis) and may be effective for chronic arthritis as well. Inhibitors of tumor necrosis factor (TNF) alpha and blockers of T cell costimulation (abatacept) are not recommended for initial therapy or systemic disease, but one of these biologic agents or methotrexate (alone or in combination with a biologic agent) can be helpful for chronic arthritis therapy [2,11]. Any biologic can be used in combination with methotrexate, another nonbiologic DMARD, and/or glucocorticoids but should not be used in combination with another biologic. Other nonbiologic DMARDs, such as cyclosporine and tacrolimus, and cytotoxic drugs, such as cyclophosphamide, are also options in patients who fail standard therapy, including biologic agents. (See 'Cyclosporine' below and 'Cytotoxic drugs' below.)

REFERRAL — Any patient who presents with moderate-to-severe sJIA, or who does not respond to nonsteroidal antiinflammatory drug (NSAID) monotherapy, should be quickly referred to an experienced pediatric rheumatologist for management, if possible.

THERAPEUTIC AGENTS — The agents used to treat sJIA include nonsteroidal anti-inflammatory drugs (NSAIDs), glucocorticoids, and the biologic and nonbiologic disease-modifying antirheumatic drugs (DMARDs).

Nonsteroidal antiinflammatory drugs — All of the traditional NSAIDs, and at least one of the cyclooxygenase (COX) 2-selective NSAIDs, have been used for treating JIA [12]. Although the traditional NSAID of first choice was aspirin, it is no longer commonly used because of dosing and toxicity issues, which warrant frequent salicylate level and liver function test monitoring. None of these NSAIDs have shown any unique advantages for children with sJIA, except possibly for indomethacin, which is anecdotally effective for control of recalcitrant fever.

If NSAID therapy alone is effective in controlling all of the sJIA manifestations, a trial of gradually tapering the NSAID can be attempted after a period of several months of disease quiescence to see if the patient has monophasic disease and has gone into remission.

Use of any of the NSAIDs is associated with a risk of hepatotoxicity, often manifested by hepatic enzyme elevation. NSAIDs are also associated with kidney toxicity, usually manifested as tubular interstitial disease with microscopic urine analysis changes, although patients can also develop edema, papillary necrosis, and hypertension. (See "Treatment and prognosis of hemophagocytic lymphohistiocytosis" and "Nonselective NSAIDs: Overview of adverse effects".)

Glucocorticoids — There are no randomized trials of glucocorticoids for the treatment of sJIA, although the Childhood Arthritis and Rheumatology Research Alliance (CARRA) consensus treatment plans and the 2013 American College of Rheumatology (ACR) guidelines offer general guidelines [2,3]. Glucocorticoids should be used judiciously in sJIA to minimize their toxicity. They are typically used during the acute phase of the illness to preserve the ability to carry out activities of daily living, particularly in children who have not responded quickly to initial therapy. (See 'Refractory acute disease' above and "Major adverse effects of systemic glucocorticoids".)

Every effort must be made to minimize the dose and duration of therapy. Whenever possible, the dose of glucocorticoids should be kept below 0.5 mg/kg per day of prednisone (or its equivalent), and the duration of therapy should be less than six months. A disease-modifying agent more appropriate for long-term management (eg, a biologic agent or methotrexate) should be initiated soon after the child is stabilized in order to facilitate the timely withdrawal of glucocorticoids. (See 'Methotrexate' below and 'Interleukin 1 inhibitors' below and 'Interleukin 6 inhibitors' below.)

Some clinicians use intermittent high-dose intravenous glucocorticoids to reduce the toxicity associated with daily oral glucocorticoids or to treat severe anemia or pericarditis. The combination of glucocorticoids and NSAIDs can result in an increased risk of gastrointestinal toxicity due to their additive effect.

Biologic disease-modifying antirheumatic drug (DMARD) therapy — sJIA is associated with increased circulating levels of multiple cytokines [13,14]. Anecdotal and experimental evidence supports the use of monoclonal antibodies or soluble receptors to block inflammatory cytokines in patients with sJIA. Of these, the most efficacious biologic agents, based upon results from randomized trials, are those that block interleukin (IL) 1 or IL-6 [15]. Many pediatric rheumatologists begin one of these agents, rather than a glucocorticoid, in patients who have failed NSAID monotherapy. (See "Overview of biologic agents in the rheumatic diseases".)

However, treatment with these particular biologics may mask some features of macrophage activation syndrome (MAS) and make it more difficult to diagnose. A systematic literature review that included 18 publications reported the clinical features of sJIA patients who developed MAS while treated with IL-1 (35 patients) or IL-6 inhibitors (49 patients) [16]. In general, patients treated with these agents had lower levels of ferritin than historical MAS cohorts. Patients treated with tocilizumab (IL-6 inhibitor) were less likely to be febrile, but other features of MAS were more pronounced, including lower platelet counts, lower fibrinogen, and higher liver function tests.

Patients receiving biologic response modifiers have an increased risk of infection, particularly mycobacterial, viral, and fungal infections. Thus, the American Academy of Pediatrics has published guidance for clinicians using these agents [17]. A thorough history is recommended to help determine infectious risk, with performance of screening tests as indicated depending upon the history and biologic agent chosen. If treatment can be safely delayed, administration of routine immunizations at least two weeks prior to starting a biologic agent is advised for inactivated or subunit vaccines and at least four weeks prior for live vaccines. Unfortunately, delaying treatment while awaiting vaccination is not feasible or safe for many children presenting with active sJIA. Administration of live vaccines is not recommended during treatment with biologic agents. An infectious disease specialist should be consulted if a live vaccine is deemed necessary while a patient is on biologic therapy. Inactivated and subunit vaccines can be given while on therapy, and an annual inactivated influenza vaccine is recommended.

There are little published data regarding withdrawal of biologic agents in children with sJIA, and no guidelines exist, although tapering and withdrawal was successful in some patients who achieved inactive disease in the tocilizumab clinical trial [18] and in an observational series of patients treated with anakinra at disease onset [4]. Most pediatric rheumatologists gradually increase the dosing interval until they feel it is safe to stop therapy in children who appear to be in remission (on drug) for six months.

The appropriate role of the various biologic agents, especially the IL-6 and IL-1 inhibitors, in the treatment of children with sJIA should become clearer as pediatric rheumatologists gain experience with these agents and their comparative effectiveness is studied [3,19]. These biologics are effective for most children with this disease, but further study is needed regarding which should be given in what situation and for what patient [20].

Interleukin 1 inhibitors — IL-1 is one of the predominant proinflammatory cytokines underlying the inflammatory symptoms of sJIA. Thus, IL-1 inhibitors such as anakinra (recombinant IL-1 receptor antagonist), rilonacept (IL-1 trap) [21], and canakinumab (anti-IL-1-beta monoclonal antibody) are beneficial in many children with sJIA. The IL-1 inhibitors, anakinra, rilonacept, and canakinumab, are administered by subcutaneous injection. Canakinumab is administered monthly and rilonacept weekly, whereas anakinra is administered daily. Because of its short half-life, the anakinra dose can be adjusted more readily or withdrawn more quickly if the patient does not respond. This is a potential advantage, especially early on in the course, when the disease is often unstable. Some features of MAS may be less pronounced in patients treated with IL-1 and IL-6 inhibitors [16]. (See 'Biologic disease-modifying antirheumatic drug (DMARD) therapy' above.)

Anakinra — Several series have demonstrated that IL-1 inhibition with anakinra is effective, although not uniformly so, in patients with both recent and established disease [4,7,22-25]. Anakinra may be more effective if used early in the disease course, rather than as "rescue" therapy once other therapies have failed. As an example, an international group gathered a case series of 46 patients who had received anakinra as part of initial therapy for sJIA, including 10 treated with anakinra without glucocorticoids or other DMARDs [7]. Many of these patients were able to avoid glucocorticoid therapy altogether, and chronic arthritis did not develop in almost 90 percent of these patients followed for more than six months (compared with 30 to 50 percent among historical controls).

In one of the subsequent prospective case series, anakinra was initiated at a dose of 2 mg/kg/day in 42 patients with new-onset sJIA, including 12 without arthritis symptoms, who had not responded to seven days of treatment with NSAIDs [26]. The dose was increased to 4 mg/kg/day if fever persisted for three days, with the option to add prednisolone or switch to another biologic agent if symptoms persisted. After one and three months of anakinra monotherapy, 55 and 71 percent of patients, respectively, had inactive disease. An additional two and five percent at one and three months, respectively, achieved inactive disease on a combination of anakinra and prednisolone. Only one patient with inactive disease on monotherapy was on the higher dose. Both patients with and without arthritis at disease onset responded to therapy. Drug tapering was begun in patients with inactive disease after a mean of 3.7 months of treatment, and 52 percent had inactive disease off medication at one year. Factors associated with inactive disease at one year included a high baseline neutrophil count (> 9x10⁹/L) and complete response after one month of therapy. Forty-two percent of patients experienced a disease flare off of therapy, with a median time to flare of five weeks. One-third of patients were switched to other biologic agents or methotrexate in combination with systemic glucocorticoids. One patient died of MAS, and three developed serious complications. This study demonstrates that treatment with IL-1 inhibition early in the course of disease is effective for many patients and may change the course of the disease. However, the response was not uniform, and some patients relapsed off treatment. In addition, since this was a single-arm open-label study, it is difficult to draw conclusions regarding the comparative efficacy of other treatments as initial therapy.

The optimal dosing of anakinra remains undefined. In one series, over 40 percent of children treated with less than 1.5 mg/kg/day required dose escalation [7]. In addition, a pharmacokinetic study done in sJIA suggested that doses between 2 to 4 mg/kg were necessary in most children with sJIA [27]. This suggests that 2 mg/kg/day is a reasonable starting dose, although some rheumatologists favor higher doses in younger children who may have different drug pharmacokinetics [3,7,25,27,28]. Subsequent dose escalation to 4 mg/kg/day or higher is necessary in some patients [3,29]. Anakinra is the first biologic of choice for many pediatric rheumatologists treating children with sJIA, although results from the first registration study are still pending. Thus, anakinra does not yet have US Food and Drug Administration (FDA) approval. However, it gained approval by the European Medicines Agency (EMA) in 2018 for use in sJIA based upon existing evidence [30].

Canakinumab — Results from two randomized trials, in addition to observational data, indicate that canakinumab (an anti-IL-1 beta monoclonal antibody) is an effective therapeutic option for patients with sJIA [9,31]. Canakinumab was approved by the US FDA for use in sJIA patients in 2013.

In the first trial, 84 children aged 2 to 19 years with active sJIA including both fever and arthritis were randomly assigned to a single dose of canakinumab (4 mg/kg subcutaneously) or placebo [9]. Concurrent treatment with another biologic agent was not allowed, but patients on background therapy of glucocorticoids, NSAIDs, and/or methotrexate were allowed to enroll. There was a significant difference in the percent of patients with an adapted JIA ACR 30 response between the canakinumab group and the placebo group (84 versus 10 percent, respectively). Two serious adverse events were reported in each group (MAS and varicella in the canakinumab group and MAS and gastroenteritis in the placebo group).

The second trial included patients from the first trial (canakinumab responders and placebo-group patients) plus additional patients (n = 100 in total) and had an initial open-label phase in which all patients were treated with canakinumab every four weeks for 12 to 32 weeks [9]. In the randomized withdrawal phase, patients who had a sustained adapted JIA ACR 30 response or better and who were not on glucocorticoids or had been tapered to a stable glucocorticoid dose were either continued on canakinumab or placed on placebo. The rate of flares was significantly lower in the canakinumab group compared with the placebo group (26 versus 75 percent, respectively). In addition, inactive disease rates were higher at the end of the withdrawal phase in the canakinumab group compared with placebo (62 versus 34 percent). Four cases of MAS were reported in the open-label phase and two in the withdrawal phase. Two patients with MAS died, one in the open-label phase and one from the placebo group in the withdrawal phase. There was no difference in the rate of serious adverse events between the two groups in the withdrawal phase.

Rilonacept — Rilonacept (an IL-1 receptor trap fusion protein) is available in the United States for treatment of autoinflammatory disorders, but its use in patients with sJIA remains investigational [32,33].

No significant differences in efficacy, based upon ACR Pediatric 30, 50, and 70 scores, were observed during the blinded phase in an exploratory trial of 24 patients with refractory sJIA randomly assigned 2:1 to rilonacept (2.2 mg/kg in cohort 1 and 4.4 mg/kg in cohort 2, given subcutaneously on days 3, 7, 14, and 21) or placebo [34]. However, fever and rash completely resolved by three months in the 23 patients who received rilonacept during the open-label phase. More than half of these patients maintained their response, and more than 90 percent were able to reduce or discontinue glucocorticoids, although 13 of the 23 patients withdrew from the study during the open-label phase before 24 months.

Additionally, a randomized trial in 71 children with active sJIA that had a four-week initial placebo-controlled phase followed by a 20-week treatment phase for all patients showed that the time to response was shorter in the rilonacept arm compared with the placebo arm. Response at week 4 was 57 percent (20 of 35 patients) in the rilonacept arm compared with 27 percent (9 of 33 patients) in the placebo arm [21]. Rilonacept was given as a loading dose of 4.4 mg/kg followed by 2.2 mg/kg weekly. Treatment was well tolerated.

Interleukin 6 inhibitors — A human recombinant monoclonal antibody against the IL-6 receptor (tocilizumab) that is administered intravenously is effective in children with severe sJIA that is resistant to conventional therapy [10,35-38]. Tocilizumab was approved by the US FDA for treatment of sJIA in 2011. Some features of MAS may be less pronounced in patients treated with IL-1 and IL-6 inhibitors [16]. (See 'Biologic disease-modifying antirheumatic drug (DMARD) therapy' above.)

In a randomized trial of tocilizumab, 112 children aged 2 to 17 years with active sJIA (arthritis, with or without fevers and other systemic symptoms) of at least six months duration who had failed previous therapy with NSAIDs and glucocorticoids were treated with placebo or tocilizumab (12 mg/kg for children <30 kg or 8 mg/kg for children ≥30 kg) intravenously every two weeks [10]. The following findings were noted:

●At 12 weeks, improvement of ACR Pedi 30, 50, and 70 responses were seen in 85, 85, and 71 percent of patients, respectively, in the treatment group compared with 24, 11, and 8 percent of patients, respectively, in the placebo group.

●Adverse events associated with tocilizumab included infections (eg, sinopulmonary infections, septic arthritis, varicella, gastroenteritis), MAS, neutropenia, and elevated liver function tests and cholesterol.

In 2018, the subcutaneous formulation of tocilizumab was approved by the US FDA for use in sJIA at the following doses: 162 mg every two weeks for patients <30 kg; 162 mg every week for patients ≥30 kg [39].

Tumor necrosis factor alpha inhibitors — In general, tumor necrosis factor (TNF) inhibitors appear to be more beneficial for children with nonsystemic categories of JIA than in those with sJIA [40,41]. This is likely due to the autoinflammatory nature of sJIA, in which innate immunity plays a larger role than in other types of JIA. Additionally, the fixed doses of etanercept that were used in the studies may be inadequate to inhibit TNF sufficiently to produce a robust response in sJIA. Other TNF inhibitors, such as infliximab, offer a wider range of dosing options, and high doses may be more effective for treating sJIA [42], but this class of biologics is not as dramatically effective as the other biologics. (See 'Interleukin 1 inhibitors' above and 'Interleukin 6 inhibitors' above.)

Several biologic agents that inhibit TNF-alpha activity are available, including the following:

●Etanercept, a soluble p75 TNF-alpha receptor fusion protein that binds TNF-alpha, has been widely tested in sJIA. However, it may provide only limited and unpredictable benefits in this disorder [40,41,43-46].

●Infliximab and adalimumab, monoclonal antibodies to TNF-alpha, also have been used with varying success in children with sJIA [41]. These agents have been effective in patients who failed to respond to etanercept, although there are limited published data.

Citing concerns about the incidence of lymphoma and other malignancies in children treated with biologic agents, the US FDA performed a safety review of malignancies reported with TNF inhibitor use. They concluded that there may be an increased risk of lymphoma and other cancers associated with the use of these drugs in children and adolescents, but this study was based solely on case reports [47]. A study published subsequently to this review that used a large administrative claims database found that children with JIA have a slightly increased risk of malignancy independent of medication usage and that medications used for JIA, including TNF inhibitors, were not significantly associated with additional malignancy risk [48]. Study of this risk is ongoing in disease-specific pharmacosurveillance registries such as the CARRA Registry, which will be able to estimate baseline compared with exposure risks for these adverse events because they will enroll all patients with JIA, regardless of medication exposure [49]. (See "Tumor necrosis factor-alpha inhibitors: Risk of malignancy".)

In addition, there may be an increased risk of reactivation of latent tuberculosis (TB) infections, as well as of fungal infections, with the use of TNF inhibitors. A US FDA alert highlighted this potential risk, especially in areas endemic for histoplasmosis, such as the Ohio and Mississippi River valleys [50]. (See 'In patients on any biologic agent' below and "Pathogenesis and clinical features of pulmonary histoplasmosis" and "Pathogenesis and clinical manifestations of disseminated histoplasmosis".)

B cell-targeted therapy — Rituximab, a B cell-depleting monoclonal anti-CD20 antibody, has been used for severe, multiple-drug-resistant sJIA with success at the case-report level [51] but has not been used with much frequency. Thus, its effectiveness remains unclear.

Other biologic DMARDs — High levels of IL-18 are correlated with severe systemic disease, MAS, and hemophagocytic lymphohistiocytosis (HLH). A recombinant IL-18 binding protein (rIL-18BP; called tadekinig alpha) is under investigation for use in adult-onset Still’s disease (AOSD) and primary forms of HLH [52]. Another potential biologic (emapalumab), which targets interferon gamma, is showing encouraging response rates in pilot study testing in patients with steroid-refractory sJIA-associated MAS [53]. Emapalumab has been approved by the FDA for use in HLH.

Nonbiologic DMARDs — The primary nonbiologic disease-modifying antirheumatic drug (DMARD) used for the treatment of sJIA is methotrexate.

Methotrexate — Methotrexate was commonly used in many children with sJIA [3,54], mostly as a steroid-sparing agent, although results from one small, randomized trial showed no difference in overall improvement in the methotrexate group compared with placebo [55]. Its use has decreased since the advent of anti-IL-1 and anti-IL-6 therapies, especially as monotherapy or in combination with glucocorticoids [3]. However, it may be effective as an adjunct to a treatment regimen containing a biologic therapy if the arthritis is not adequately controlled.

Hepatic toxicity may occur with methotrexate therapy but does not seem to be any more frequent than expected with the use of methotrexate in other subtypes of JIA. There are a number of other common toxicities seen with methotrexate therapy and several potential serious adverse effects. These are discussed in greater detail separately. (See "Hepatotoxicity associated with chronic low-dose methotrexate for nonmalignant disease" and "Major side effects of low-dose methotrexate".)

The general range in which methotrexate is prescribed for JIA is an oral dose of up to 25 to 30 mg/m² (0.5 to 1 mg/kg) per week, with a maximum dose of 25 mg per week. Above 15 to 20 mg/m² (roughly 0.5 mg/kg), oral absorption is unreliable, and parenteral administration may be advantageous [56]. Folic acid or leucovorin (folinic acid) should be given to minimize the gastrointestinal side effects associated with methotrexate therapy. (See "Polyarticular juvenile idiopathic arthritis: Treatment", section on 'Methotrexate'.)

Thalidomide — Thalidomide is a unique agent that is effective in both suppressing the production of cytokines such as TNF-alpha and IL-6 [57,58] and in blocking angiogenesis [59]. Case reports and small series suggest that thalidomide is an effective medication for children with sJIA [60]. With the advent of effective IL-1 and IL-6 inhibitors, however, thalidomide and its derivatives should be reserved for children in whom the biologics are not options due to cost and availability. (See 'Interleukin 1 inhibitors' above and 'Interleukin 6 inhibitors' above.)

Neuropathy and sedative effects are often dose limiting. This agent is an infamous teratogen, and clinical use is strictly regulated. Effective contraception and monitoring for pregnancy are prerequisites for use in girls of childbearing age. Despite these cautions, it has been dramatically effective for some children who have failed other therapies. Thalidomide may be especially important in countries where biologic agents are not available because of cost issues. Revlimid (lenalidomide) is an alternative agent related to thalidomide that may have a lower incidence of side effects, but there are no published trials of its use in sJIA.

Leflunomide — Leflunomide is an isoxazole derivative that has effects on many components of immune and inflammatory responses. In case reports and small series, leflunomide has been used successfully to treat some patients with poor response to, or side effects from, methotrexate [61,62]. However, leflunomide was less effective than methotrexate and had a similar side-effect profile in one head-to-head randomized trial of 94 patients with polyarticular JIA [63]. This agent has not been studied specifically in sJIA.

Cyclosporine — Cyclosporine (as well as tacrolimus, another calcineurin inhibitor) has been used with some frequency in sJIA, mainly in patients with acute or chronic MAS and chronic refractory systemic activity [64,65]. (See "Systemic juvenile idiopathic arthritis: Course, prognosis, and complications" and "Treatment and prognosis of hemophagocytic lymphohistiocytosis".)

Cytotoxic drugs — Oral and intravenous cyclophosphamide, etoposide, chlorambucil, and azathioprine have been used in the treatment of sJIA with varying success. Cyclophosphamide and etoposide are the only cytotoxic drugs used with any frequency. Cyclophosphamide has been used for a small subset of children with severe refractory sJIA, generally prior to the availability of biologic agents; however, its use was associated with sustained improvement in these patients in a small study [66].

Cytotoxic medications are generally reserved for children who have progressive disease despite extensive use of less toxic medications, such as the biologic therapies, particularly the IL-1 and IL-6 inhibitors. They have also been used for complications of sJIA, including MAS, pulmonary hypertension, and interstitial lung disease. Leukopenia, infections, the risk of sterility, and the risk of subsequent neoplastic disease limit their utility. (See "General toxicity of cyclophosphamide in rheumatic diseases" and 'Biologic disease-modifying antirheumatic drug (DMARD) therapy' above and "Systemic juvenile idiopathic arthritis: Course, prognosis, and complications", section on 'Macrophage activation syndrome'.)

Oral small molecule enzyme inhibitors

Janus kinase inhibitors — Janus kinase (JAK) inhibitors (such as tofacitinib) are under investigation in clinical trials of patients with sJIA. There are anecdotal reports of potential efficacy in case reports [67].

HEMATOPOIETIC STEM CELL TRANSPLANTATION — Hematopoietic stem cell transplantation (HCT) has been used in children with relentless disease that is uncontrolled despite the use of multiple drugs [68,69]. However, most children were treated with HCT prior to the advent of effective biologic therapies. There is some continued interest in HCT for severe resistant sJIA [68,69], but the procedure still carries a high mortality rate [70]. Thus, HCT should be restricted to only the most severely affected children refractory to other, less risky therapies.

An observational study of autologous stem cell transplantation included 34 children treated at several European centers, 29 of whom had sJIA and 5 with severe polyarticular JIA [71]. At follow-up that ranged from 12 to 60 months, 18 of 34 (53 percent) were in drug-free remission, 6 improved, and 7 had no benefit. Three children died due to macrophage activation syndrome (MAS; secondary hemophagocytic lymphohistiocytosis [HLH]) that may have been infection related. In a subsequent report that involved many of the same subjects over a median follow-up period of 80 months, two died early in follow-up from MAS, and, of the remaining 20 patients, 8 were in drug-free remission, 7 were in partial remission, and 5 had no benefit [72]. Of these latter five patients, two patients died from infection. (See "Treatment and prognosis of hemophagocytic lymphohistiocytosis".)

A subsequent series reported on 16 JIA patients, including 11 with sJIA, who were treated with reduced-intensity conditioning and allogeneic HCT [73]. One of the patients with sJIA had significant graft-versus-host disease and died of pulmonary hemorrhage, but all remaining patients had improvement in their disease, with 8 of the surviving 10 patients with sJIA in complete remission off medications at their last follow-up visit, making HCT a potential option in patients with severe and/or refractory disease.

ADDITIONAL MANAGEMENT ISSUES

Macrophage activation syndrome — There is an anecdotal association of initiation of any therapy in sJIA, including nonsteroidal antiinflammatory drugs (NSAIDs), with a transiently elevated risk of life-threatening macrophage activation syndrome (MAS), a form of hemophagocytic lymphohistiocytosis (HLH). Accordingly, careful monitoring for MAS after initiation of any therapy for sJIA is essential. Patients with sJIA are predisposed to developing MAS and may display features of MAS throughout their disease course, with some patients having evidence of "chronic MAS" that is refractory to treatment and carries a poor prognosis. Treatments for HLH and MAS are under investigation in clinical trials (see 'Other biologic DMARDs' above). Diagnosis and treatment of MAS are discussed in greater detail separately. (See "Systemic juvenile idiopathic arthritis: Course, prognosis, and complications", section on 'Macrophage activation syndrome' and "Clinical features and diagnosis of hemophagocytic lymphohistiocytosis", section on 'Clinical features' and "Treatment and prognosis of hemophagocytic lymphohistiocytosis", section on 'MAS/Rheumatologic conditions'.)

Other complications — Management of other complications related to sJIA (eg, pulmonary complications) and systemic glucocorticoid treatment (eg, growth retardation, osteoporosis, iron-deficiency anemia) are discussed in detail separately. (See "Systemic juvenile idiopathic arthritis: Course, prognosis, and complications", section on 'Complications' and "Systemic juvenile idiopathic arthritis: Clinical manifestations and diagnosis", section on 'Laboratory findings'.)

Monitoring/surveillance — Studies are required to monitor for drug toxicities for several of the therapies used to treat sJIA. Patients with sJIA can be quite unstable, especially during the first few months of disease, and need frequent monitoring of systemic disease activity and potential complications such as MAS or pulmonary disease, aside from what is required because of specific medication toxicities. All sJIA patients should have at least the following blood tests done regularly: complete blood count, aspartate transaminase (AST), alanine transaminase (ALT), erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and ferritin. (See "Systemic juvenile idiopathic arthritis: Course, prognosis, and complications", section on 'Complications'.)

In patients on nonsteroidal antiinflammatory drugs — Careful monitoring for hepatic and renal toxicity is required if NSAIDs are used. This includes complete blood counts, kidney and liver function tests, and urine analyses within several weeks of starting the medication and then every three to four months if the patient remains on a stable dose. (See 'Nonsteroidal antiinflammatory drugs' above and "Nonselective NSAIDs: Overview of adverse effects".)

In patients on any biologic agent — Use of biologic therapy, particularly the tumor necrosis factor (TNF) alpha inhibitors, is associated with reactivation of latent tuberculosis (TB) infection. Thus, screening for TB is recommended prior to starting a biologic agent. Annual rescreening is not necessary in low-risk children [2]. However, repeat screening is advised if the risk of TB increases to moderate or high. (See "Risk of mycobacterial infection associated with biologic agents and JAK inhibitors" and "Tuberculosis infection (latent tuberculosis) in children", section on 'Diagnosis'.)

In patients on tocilizumab — Recommended laboratory monitoring in patients on tocilizumab, a human recombinant monoclonal antibody against the interleukin (IL) 6 receptor, includes neutrophil and platelet counts, liver enzyme (AST/ALT) levels, and lipid levels at least monthly until stable and then every three months (six months for lipid levels). (See 'Interleukin 6 inhibitors' above.)

In patients on TNF-alpha inhibitors — Patients on tumor necrosis factor (TNF) alpha inhibitors are at increased risk for fungal infections, as well as dissemination of TB. Patients should be screened for TB exposure prior to receiving TNF-alpha inhibitors and annually if they are at high risk of exposure. In addition, they should be monitored for histoplasmosis in endemic areas. (See "Clinical manifestations, diagnosis, and treatment of miliary tuberculosis" and "Pathogenesis and clinical manifestations of disseminated histoplasmosis" and "Diagnosis and treatment of disseminated histoplasmosis in patients without HIV" and "Tuberculosis infection (latent tuberculosis) in adults: Approach to diagnosis (screening)".)

In patients on methotrexate — Appropriate clinical and laboratory monitoring is essential for patients treated with methotrexate. Laboratory monitoring includes aminotransaminases, albumin, and complete blood count every six to eight weeks, until a stable dose is reached, and then every three months if all laboratory testing is normal. Toxicities and monitoring are discussed in greater detail separately. (See "Hepatotoxicity associated with chronic low-dose methotrexate for nonmalignant disease" and "Major side effects of low-dose methotrexate".)

Immunizations — The administration of live-viral vaccines and other standard childhood immunizations in patients with JIA is discussed in detail separately. (See "Oligoarticular juvenile idiopathic arthritis", section on 'Immunizations'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Juvenile idiopathic arthritis".)

SUMMARY AND RECOMMENDATIONS

●Initial treatment for mild-to-moderate symptoms – In children with mild-to-moderate, nondisabling symptoms, we suggest a nonsteroidal antiinflammatory drug (NSAID) other than aspirin as initial therapy for systemic juvenile idiopathic arthritis (sJIA) (Grade 2C). In general, a trial of NSAIDs alone should last no more than a few weeks. An additional agent should be added in children who develop or continue to have significant symptoms despite use of NSAIDs. (See 'Mild-to-moderate acute disease' above and 'Nonsteroidal antiinflammatory drugs' above.)

●Rationale for choice of a glucocorticoid or biologic – Historically, most pediatric rheumatologists have added a glucocorticoid in patients who have not responded to a trial of an NSAID alone or who have severe disease on initial presentation. However, long-term use of glucocorticoids is associated with a high frequency of side effects. Biologic agents, especially the interleukin (IL) 1 and IL-6 inhibitors, are being used more and more as single therapy at disease onset because they are effective in reducing clinical symptoms in patients with disease refractory to NSAID and glucocorticoid therapy. There are increasing data to suggest that they are also beneficial when used early in the care of children with severe sJIA (ie, instead of glucocorticoids in selected patients) once the diagnosis is confirmed. The decision to initiate therapy with a biologic agent first or in combination with glucocorticoids and the choice of biologic agent must be made by the individual treating clinician after discussion of the risks and benefits with the patient and family. (See 'Moderate-to-severe acute disease' above and 'Glucocorticoids' above and 'Biologic disease-modifying antirheumatic drug (DMARD) therapy' above.)

●Initial treatment of severe symptoms and for patients in whom NSAIDs were ineffective – For patients in whom a nonsteroidal antiinflammatory drug (NSAID) alone was not effective, or whose initial symptoms include high fevers, other severe systemic symptoms, and/or disabling polyarthritis, we suggest adding one of the biologic agents, such as anakinra, canakinumab, or tocilizumab, rather than or in addition to glucocorticoids (Grade 2C). Anakinra is initially prescribed at a dose of 2 mg/kg daily (up to 100 mg). Canakinumab is usually given as 4 mg/kg monthly in children (maximum dose 300 mg). Tocilizumab is typically given every two weeks at a dose of 12 mg/kg intravenously or 162 mg subcutaneously for children who weigh less than 30 kg and 8 mg/kg intravenously (up to 800 mg) every two weeks or 162 mg subcutaneously every week for those who weigh more. (See 'Moderate-to-severe acute disease' above and 'Interleukin 1 inhibitors' above and 'Interleukin 6 inhibitors' above.)

●Refractory disease – We suggest adding a glucocorticoid if there is not a prompt response to a biologic agent (Grade 2C). While the definition of a prompt response will differ from center to center, few pediatric rheumatologists would wait more than a week before adding a glucocorticoid in the face of continued polyarthritis, fever, and rash, and all would add a glucocorticoid sooner if there was evidence of macrophage activation syndrome (MAS) or severe serositis. Both the glucocorticoid and the biologic agent are usually continued until disease control is established. The clinician may then decide to gradually withdraw the glucocorticoid first because of the inevitable toxicity associated with its chronic use. Withdrawal of the biologic agent may be attempted as well if the patient continues to exhibit no sign of disease. The prednisone dose should be limited to 0.5 to 1 mg/kg, if possible, although doses as high as 2 mg/kg, or pulsed high-dose therapy, may be required in severe cases. (See 'Refractory acute disease' above and 'Glucocorticoids' above.)

●Recurrent or polyphasic disease – The approach to treating patients with a polyphasic disease course and recurrent episodes is the same as for initial therapy. (See 'Recurrent or polyphasic disease' above and 'Initial therapy' above.)

●Persistent chronic disease – Treatment of persistent chronic disease depends upon whether the patient has primarily systemic manifestations (eg, fever, rash, and serositis), primarily arthritis (which can be progressive and destructive), or both types of manifestations. Inhibitors of IL-1 and IL-6 appear to be the most effective of the biologic agents for primarily systemic disease and may be effective for chronic arthritis as well. Inhibitors of tumor necrosis factor (TNF) alpha, blockers of T cell costimulation (abatacept), and methotrexate can be helpful adjuncts for chronic arthritis therapy [2,11]. Other nonbiologic disease-modifying antirheumatic drugs (DMARDs), such as cyclosporine and tacrolimus, and cytotoxic drugs, such as cyclophosphamide, are additional options in patients who fail standard therapy, including biologic agents.

●When to refer – The potential toxicities of the drugs used to treat sJIA, including MAS, must be carefully balanced against the aggressive and often permanently disabling nature of unremitting disease. Thus, patients who present with severe sJIA or who are refractory to NSAIDs should be referred to an experienced pediatric rheumatologist for management, if possible. (See 'Overview' above and 'Referral' above.)

●Monitoring disease activity and surveillance for complications and medication toxicities – Patients need frequent monitoring of systemic disease activity and surveillance for potential complications such as MAS or pulmonary disease. Studies for potential medication toxicities are required for several of the therapies used to treat sJIA. (See 'Monitoring/surveillance' above.)

●Management of complications – The management of complications associated with sJIA, including MAS, are discussed in detail separately. (See "Systemic juvenile idiopathic arthritis: Course, prognosis, and complications", section on 'Complications'.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Thomas JA Lehman, MD, who contributed to earlier versions of this topic review.