Interfering with interferons: targeting the JAK-STAT pathway in complications of systemic juvenile idiopathic arthritis (SJIA)

Abstract

Systemic JIA (SJIA) is distinguished from other forms of JIA by the prevalence of the severe, life-threatening complications macrophage activation syndrome (SJIA-MAS) and lung disease (SJIA-LD). Alternative therapeutics are urgently needed, as disease pathogenesis diverges from what is observed in SJIA, and currently available biologics are insufficient. SJIA-MAS, defined by a cytokine storm and dysregulated proliferation of T-lymphocytes, and SJIA-LD which presents with lymphocytic interstitial inflammation and pulmonary alveolar proteinosis, are both thought to be driven by IFNs, in particular the type II IFN-γ. Involvement of IFNs and a possible crosstalk of type I IFNs with existing biologics indicate a distinct role for the JAK-STAT signalling pathway in the pathogenesis of SJIA-MAS and SJIA-LD. Here, we review this role of JAK-STATs and IFNs in SJIA complications and discuss how new insights of ongoing research are shaping future therapeutic advances in the form of JAK inhibitors and antibodies targeting IFNs.

Rheumatology key messages.

• Available biologics are insufficient to treat the severe complications SJIA-MAS and SJIA-LD.

• Interferons, signaling via the JAK-STAT pathway, play a major role in SJIA-MAS and SJIA-LD.

• Targeting interferons and the JAK-STAT pathway may shape the next generation of therapeutics for SJIA-MAS/LD.

SJIA and current therapies

JIA describes a group of heterogeneous childhood-onset diseases with unknown aetiology. The different JIA disease subtypes have diverging pathophysiologic origins and mechanisms, ranging from various degrees of adaptive and innate immune dysfunction, production of autoantibodies, and dysregulation of immune cell populations such as T cells and monocytes. However, all JIA subtypes are unified by the presence of chronic childhood arthritis [1, 2]. In systemic JIA (SJIA), arthritis can sometimes play a minor role at disease onset and instead systemic inflammation is predominant, with symptoms including fever, rash, lymphadenopathy, hepatosplenomegaly and serositis [3]. Systemic JIA is further set apart from other forms of JIA by its prevalence for the serious complications macrophage activation syndrome (SJIA-MAS) and lung disease (SJIA-LD) [4, 5].

Treatments for pediatric rheumatologic diseases such as SJIA have significantly evolved in the past two decades. Synthetic glucocorticoids that suppress inflammation were developed >60 years ago, and have since been widely used in various chronic diseases, including SJIA [6]. DMARDs such as methotrexate were the first non-steroidal medications shown to significantly improve arthritis in JIA, and together with NSAIDs and glucocorticoids formed the mainstay of treatment in the pre-biologic era [7, 8]. However, methotrexate proved largely ineffective for the systemic features of SJIA, and adverse events, including steroid side effects, had significant impact on pediatric patients [6, 9].

With better understanding of specific inflammatory cytokines and proteins in SJIA disease pathology, including IL-1, IL-6, TNFα, IL-18 and S100 proteins [10], biologic agents in the form of recombinant monoclonal antibodies or recombinant proteins, directed at single cytokines, were developed over the past two decades. The first biologic approved for clinical use in JIA was etanercept in 1999, which binds to circulating TNFα and prevents its interaction with cell surface receptors [11]. Other biologics have since been developed and used and/or approved specifically for SJIA, including anakinra (a recombinant IL-1 receptor antagonist, approved by the European Medicines Agency (EMA) in 2018) and canakinumab (a recombinant monoclonal IL-1β antibody, approved by the EMA and US Food and Drug Administration (FDA) in 2013), and tocilizumab (a recombinant antibody against IL-6, approved by the EMA and FDA in 2011) [7, 12]. Since biologic drugs have been introduced in the clinical practice and management of JIA, the prognosis for pediatric patients has dramatically improved [13–15].

Despite this success, between 20% and 40% of SJIA patients fail to respond to anti-cytokine biologics or develop adverse events during treatment [16, 17]. Some treatments can also lose efficacy over time, resulting in the need for novel therapeutic strategies. For example, in the randomized controlled trial of canakinumab, 38% of SJIA patients do not achieve an adapted JIA ACR criteria (JIA-ACR) score of at least 50 [18]. Additionally, the potentially life-threatening complications SJIA-MAS and SJIA-LD are not prevented by currently available therapies, as they instead seem to be driven by IFNs and IL-18 (discussed below) [4, 5, 19]. While individual pro-inflammatory cytokines present an obvious target in many rheumatologic diseases, evidence has accumulated for the significance of the Janus kinase (JAK)-signal transducer and activator of transcription (STAT) signalling pathway in disease pathogenesis. In particular, the key role of this pathway in IFN signalling highlights the importance of JAK-STAT signalling as a potential therapeutical target for SJIA and its complications.

Interferons and JAK-STAT signalling

The JAKs compose a family of four intracellular tyrosine kinases JAK1, JAK2, JAK3 and tyrosine kinase (TYK) 2. When cytokines bind to membrane receptors, JAKs are activated, recruited and in turn phosphorylate the receptors. This allows the selective binding and phosphorylation of members of the STAT family to induce downstream genes transcription [6]. Seven STATs have been identified in mammals: STAT1, STAT2, STAT3, STAT4, STAT5A, STAT5B and STAT6 [20].

Overall, >50 cytokines signal via the JAK-STAT pathway, including IL-2, IL-6 and IL-10 family members, to regulate cell homeostasis, proliferation and differentiation, as well as control the immune system and inflammatory response. However, the JAK-STAT pathway also plays a central role in mediating the cellular response to the IFN family cytokines.

IFN type I, consisting of 13 subtypes of IFNα and one IFNβ, binds to the IFN alpha receptor (IFNAR) 1 and IFNAR2 heterodimer, signalling downstream via JAK1 and TYK2. In response, STAT1 and STAT2 are phosphorylated, heterodimerize, and are transported into the nucleus, where they then associate with IFN regulatory factor (IRF) 9 to form a transcriptionally active complex termed the IFN stimulatory gene factor (ISGF) 3 (Fig. 1). ISGF3 then recognizes specific sequences (IFN-stimulated response elements, ISRE) in IFN-stimulated genes (ISGs), where it binds and facilates their transcription [21].

Fig. 1.

Interferon-induced JAK-STAT signalling pathways

IFN-α/β, IFN-γ and IFNλ signal via distinct receptors, but can share downstream JAK-STAT pathways and signalling molecules. Canonically, IFN-α/β and IFNλ induce assembly of the interferon-stimulated gene factor 3 (ISGF3) which promotes transcription by recognizing and binding to interferon-stimulated response elements (ISREs). IFN-γ canonically leads to the assembly of STAT1 homodimers, which migrate to the nucleus and bind to interferon gamma activation sites (GASs), inducing transcription of ISGs. Unphosphorylated STATs (U-STAT) are also able to facilate transcription via GAS.

Type II IFN, IFN-γ, binds to the IFN gamma receptor (IFNGR) heterodimer IFNGR1/IFNGR2 and utilizes JAK1 and JAK2 to induce phosphorylation of STAT1. STAT1 forms homodimers, which in the nucleus bind to Gamma-IFN sites (GAS) to activate transcription of ISGs (Fig. 1). Finally, type III IFNs, composed of IFNλ1-3, use the same signalling pathway as type II IFN with the exception of binding to a heterodimeric receptor of IL-28Rα and IL10Rβ [21].

While the general mode of IFN-mediated JAK-STAT signal transduction is known, various forms of non-canonical signalling have been discovered, along with different combinations of STAT homo- and heterodimers to convey potentially very different transcription signals. For example, type I IFNs are also able to induce homodimers of STAT1 or STAT3 instead of utilizing the ISGF3 for transcriptional signalling. These IFN type I induced STAT homodimers bind to GAS to induce gene expression [22]. Curiously, while type I and type III IFN both utilize the ISGF3, their ISGF3-induced gene expression profile is not identical [23]. This indicates that mere activation of the same transcriptionally active complexes does not convey which genes are expressed, suggesting that either IFN dosage or presence of other factors can divert ISGF3 to other loci [24]. Type I IFNs also induce responses by different subtypes alone, and patterns of STAT activation can vary depending on cell types [25]. To complicate the matter, unphosphorylated STATs can play a role in nuclear gene expression. Unphosphorylated STAT1 and 3 (U-STAT1, U-STAT-3) (Fig. 1), for example, can enter the nucleus and modulate gene transcription by binding to DNA or to transcription factors such as NFκB [26].

Assessment of the IFN response in disease

To date, it remains difficult to distinguish the specific roles of type I and type II IFN. Direct serum IFN measurements using ELISA have proven to be challenging due to the presence of heterophilic serum proteins that nonspecifically bind to the capture and detection antibodies in the assay, as well as extremely low biological levels of type I IFN in serum [27, 28]. Only recently have new assays been developed that can detect attomolar, physiological levels of IFNα [28]. Therefore, the measuring of IFN-induced chemokines CXCL9, CXCL10 and CXCL11 is widely accepted. CXCL9 is induced by IFN-γ but not IFN-α or -β [29], CXCL10 is induced by IFN-γ and to a lesser degree by IFN-α/β [30] and CXCL11 is similarly induced by IFN-γ and IFN-β, and weakly by IFN-α [31]. Notably, only CXCL9 is completely dependent on IFN-γ and can thus be used as secondary measurement for presence of type II IFN [32]. CXCL10, CXCL11 and neopterin, however, are not selectively induced by IFN-γ despite often being reported as such [33, 34]. Together, this may result in a certain bias towards the role of IFN-γ, while IFN-α and IFN-β responses are not properly assessed. In addition, IFN gene score signatures are used to evaluate the role of IFNs in diseases by assessing the expression of ISGs compared with healthy individuals, and have particularly been utilized in interferonopathies or IFN driven diseases such as systemic lupus erythematosus (SLE). Assessing the IFN gene score by RNA level is a more sensitive method to pick up IFN activity as compared with protein levels, particularly when compared with direct measurement of IFNs. Distinct IFN gene signatures have been developed for type I and type II IFNs, though overlapping transcriptional activation as well as a dynamic IFN signatures that evolve over time continue to make interpretation challenging [35–38].

Role of IFNs and JAK-STAT in SJIA

Polyarticular and oligoarticular JIA have been associated with high plasma levels of IFN-γ, as well as elevated CXCL9/10 in both plasma and synovial fluid [39].

However, SJIA (when not complicated by MAS or LD) is not generally associated with a robust type I or type II IFN signature. Independent microarray gene expression studies using PBMCs did not detect any IFN-induced signature in SJIA [40–42], with the exception of one study on a Japanese SJIA patient cohort [43]. Recent bulk RNA-Seq analysis of SJIA patient monocytes also failed to reveal evidence of an IFN-induced signature [44].

No recent studies have reported detectable type I IFN levels in serum of SJIA patients, owing to the lack of specific, highly sensitive methods to detect low IFN-α/β in serum [27, 28, 45].

In contrast, reports on IFN-γ and CXCL9/10 in SJIA have been contradictory. Gattorno et al. reported that CXCL10 and IFN-γ were modestly elevated compared with controls but did not display the control levels. De Jager et al. also showed that while IFN-γ was not elevated in SJIA patients with longstanding disease, CXCL9 and CXCL10 were, albeit at significantly lower levels than those seen in oligoarticular or polyarticular JIA [39, 46]. Similarly, a recent study reported relatively low but increased CXCL9 levels in SJIA patients that respond to canakinumab vs non-responders [47]. On the other hand, more recent studies have not observed significantly elevated IFN-γ in SJIA patients’ serum [39, 45], or increased IFN-γ production following SJIA PBMC stimulation compared with healthy controls [48]. Similarly, Bracaglia et al. did not observe significantly elevated levels of IFN-γ or IFN-γ inducible chemokines CXCL9, CXCL10 or CXCL11 in active SJIA patients without MAS [49]. SJIA patients’ natural killer (NK) cells in fact have been shown to have a specific defect in IL18-induced IFN-γ production, at least in part caused by a defective phosphorylation of the IL 18 receptor beta [50, 51].

Intriguingly, however, a subset of SJIA patients without MAS treated with anti-IL-1 inhibitors were shown to upregulate a type I IFN gene signature [37, 52]. Monocytes from SJIA patients naïve to biological therapy are hyporesponsive to IFN-γ and lack an IFN-induced gene expression signature, but anti-IL1b treatment can increase their basal IFN signal and also increase monocyte responsiveness to IFN-γ, potentially facilitated by increased levels of IFNGR expression [44, 53, 54].

Role of IFNs and JAK-STAT in the complications SJIA-MAS and SJIA-LD

In contrast to SJIA without MAS, substantial evidence supports a key role for type II IFNs in SJIA-MAS. MAS is a severe and potentially fatal hyperinflammatory complication arising in up to 30% of SJIA patients [5, 55]. Key features of MAS are overactivation of T lymphocytes causing highly elevated production of IFN-γ, which drives activation of hemophagocytic macrophages. What rheumatologists call MAS is also called secondary hemophagocytic lymphohistiocytosis (HLH) when resulting after infection or other triggers [56]. Primary HLH on the other hand is an autosomal recessive disorder caused by deficiencies in genes involved in the cytolytic activity, such as perforin, resulting in impaired NK cell and cytotoxic T-cell function and inability to limit proliferation and expansion of T cells and macrophages [55, 57]. While the pathophysiology is not completely understood, it is thought that risk for MAS is at least partially driven by a genetic component, and indeed several studies have found SJIA-MAS patients carrying heterozygous variants in causative genes for primary HLH [58–60].

Because IFN-γ production plays a key role in MAS, high levels of IFN-γ and IFN-γ inducible chemokines CXCL9 and CXCL10 in serum are characteristic for MAS patients [45, 49, 61]. Bracaglia et al. found that IFN-γ and CXCL9, CXCL10 and CXCL11 are highly elevated in SJIA-MAS patients [49], and another study showed MAS is characterized by IFN-γ production from CD8+ lymphocytes [61]. Murine MAS models have further increased evidence for the pivotal role of IFN-γ in MAS [62–65]. Recently, single-cell RNA-sequencing of bone marrow macrophages in early SJIA-MAS revealed activated subpopulations with altered transcriptomes including upregulated IFN-γ response pathways [44].

IL-18, known for its strong IFN-γ inducing capacities, is also highly increased in SJIA-MAS [36, 47, 49]. Hyperproduction of IL-18 drives the overactivation of Th1 lymphocytes and macrophages. Massive hypersecretion of IFN-γ from these cells is in fact likely a result of high IL-18 levels setting the milieu for MAS [10, 36, 45]. In a murine MAS model, mice lacking the natural inhibitor of IL-18, IL-18 binding protein, had a more severe clinical manifestation of MAS compared with wildtype mice, highlighting the critical role for this protein in driving disease pathogenesis [62, 66].

Another SJIA-associated complication with links to IL-18 and IFN activation is interstitial lung disease (SJIA-LD) [4, 19]. SJIA-LD is characterized by lymphocytic interstitial inflammation as well as accumulation of lipoproteinaceous material and lipid-laden macrophages in the lungs. These are features shared with pulmonary alveolar proteinosis (PAP), where dysfunction of alveolar macrophages by lack of GM-CSF signalling leads to accumulation of pulmonary surfactant in alveolar spaces [4, 67]. In SJIA-LD, GM-CSF signalling remains intact, but substantial Th1-driven lung inflammation is present, along with highly increased IL-18 serum levels, IL-18 and IFN-γ-induced chemokines in the lungs, and pulmonary gene expression reflecting IFN-driven activation [4].

The majority of patients with SJIA-LD were diagnosed after 2000, have refractory systemic disease with MAS, and are being treated with various biological therapies. Saper and colleagues have hypothesized that the timeline of increase in SJIA-LD aligning with the rise of cytokine-directed therapies may indicate a connection of both [19]. The underlying pathogenesis of SJIA-LD remains unknown, and considerations include persistently active SJIA-MAS perhaps modified by biologic therapy, a delayed hypersensitivity reaction to anti-IL1 or anti-IL-6 agents, or an autoinflammatory reaction towards cryptic antigens exposed by persistent inflammation during SJIA-MAS [68, 69].

Therapeutic approaches targeting IFNs in SJIA, SJIA-MAS and SJIA-LD

While IFN-γ appears activated in polyarticular and oligoarticular JIA, it is likely not the main driver of disease. However, in SJIA-MAS, SJIA-LD and HLH, therapy directed at IFN-γ appears much more promising. As noted above, MAS mouse models using repeated CpG DNA induced TLR9 stimulation have shown that direct IFN-γ blockade or JAK/STAT inhibition alleviates inflammation [62, 64,65, 70]. Another MAS mouse model using animals transgenic for human IL-6 challenged by LPS provided evidence that direct IFN-γ inhibition with monoclonal antibodies may be effective in SJIA-MAS [65].

Emapalumab is a monoclonal antibody against human IFN-γ. A recent open-label phase 2–3 study evaluating the efficacy and safety of emapalumab in children with HLH found the treatment efficacious without any organ toxicity, although infections, including severe infections, occurred frequently [32]. Thus far, emapalumab treatment for SJIA-MAS is based on anecdotal clinical experience; an open-label clinical trial is ongoing (Clinicaltrials.gov: NCT03311854), though the preliminary data is promising [71]. There are case reports of emapalumab in secondary HLH/MAS, including a young adult patient with adult-onset Still’s disease and MAS successfully treated with emapalumab following seven infusions within 2 months [72]. Emapalumab was also effective in treating a patient with refractory Epstein–Barr (EBV)-associated secondary HLH despite severe pre-existing comorbidities [73]. However, it remains to be seen if neutralization of IFN-γ alone by emapalumab is an effective therapy outside of HLH, where it was mainly utilized as a bridge to bone marrow transplant [32, 74].

Inhibition of type I IFN has not been investigated in clinical trials for JIA or SJIA patients, as there is no direct evidence for IFN type I contribution to the pathophysiology at this time. Therapy targeting type I IFN, however, has been investigated in other rheumatic diseases. Studies of sifalimumab, an anti-IFN alpha monoclonal antibody, were discontinued despite promising phase 2 b results in systemic lupus erythematosus (SLE), an autoinflammatory disease marked by high elevation of type I IFN gene signatures [75]. Instead, anifrolumab, a monoclonal antibody directed against type I IFN receptor subunit 1 that has recently been found effective in a phase 3 trial in SLE is now FDA approved [76]. Should future studies be able to overcome the difficulties of disentangling type I and II IFN effects and provide evidence of IFN type I driven immunity in JIA or particularly in SJIA-MAS and SJIA-LD, alternative therapies such as anifrolumab should be considered.

Indirect inhibition of IFNs: JAK inhibitors in SJIA, SJIA-MAS and SJIA-LD

In contrast to recombinant antibodies directly targeting IFNs or their receptors, much more is known about the effectiveness of JAK inhibitors (jakinibs), a novel group of oral small molecule inhibitors.

Jakinibs target both type I and type II IFN (and type III, though not further discussed here) pathways, as well as other cytokines, by interfering with the JAK-STAT pathway. JAKs provide high biological plausibility as efficacious targets in diseases where either IFNs, other cytokines signalling via the JAK-STAT pathway, or both are driving the disease. These small molecule inhibitors use a novel mechanism of action by affecting intracellular signalling pathways instead of targeting a specific cytokine or its receptor. The competitive interaction of the Jakinib with the JAK region constituting the ATP binding site significantly interferes with JAK and STAT phosphorylation required for downstream signalling [77]. Various jakinibs with different selectivity towards JAK1, JAK2, JAK3 and TYK2 have been developed, though the homology within the JAK family structures makes it challenging to target just one specific JAK selectively [77]. Ruxolitinib and baricitinib are more selective towards JAK1 and JAK2 over the other JAKs. Tofacitinib, on the other hand is most effective at blocking JAK3 [77, 78]. A recent study found that tofacitinib, baricitinib, and two other mainly JAK3 inhibitors named upadacitinib and filgotinib exhibit similar cytokine receptor inhibition profiles in vitro despite different IC₅₀ values for each JAK [79]. Fenwick et al. tested two different JAK inhibitor compounds and their effectiveness at inhibiting CXCL9, CXCL10 and CXCL11 secretion from chronically obstructive pulmonary disease (COPD) patients’ airway epithelial cells. They found both compounds suppressed the chemokines and STAT1 phosphorylation, but one compound (PF1367550) was more effective than the other [80]. Thus, whether subtle differences in structure and selectivity could translate to clinical differences between these JAK inhibitors remains to be seen.

While the Jakinib tofacitinib has been recently FDA approved for treatment of polyarticular course JIA, and there is an ongoing clinical trial for SJIA with systemic features (Clinicaltrials.gov: NCT03000439), currently only case reports support these medications for SJIA or its complications, MAS and SJIA-LD. In fact, use of jakinibs in SJIA-MAS and SJIA-LD is extrapolated from murine data and from clinical trials with similar diseases that are driven by IFN.

Jakinibs have been assessed in monogenic interferonopathies such as SAVI [stimulator of IFN genes-associated (STING-associated) vasculopathy with onset in infancy] and CANDLE (chronic atypical neutrophilic dermatosis with lipodystrophy and elevated temperatures), with results showing clinical and laboratory improvement of disease [81, 82]. As noted above, murine models have also shown beneficial effects of JAK/STAT inhibitors in HLH and MAS. Ruxolitinib significantly reduced the clinical and laboratory manifestations of both primary HLH (PRF^-/- mice infected with LCMV) and MAS/secondary HLH (mice repeatedly stimulated with CpG DNA) [64, 70, 83].

One report describes using ruxolitinib in an EBV-induced secondary HLH patient, resulting in decreased disease markers, including ferritin. However, with worsening clinical status, treatment was ceased after only 7 days and the patient expired [84]. On the other hand, a pilot study of a 28-day treatment course of oral ruxolitinib in 12 children with secondary HLH showed encouraging results, with two-thirds of patients achieving complete response and maintaining this status for >6 months [85].

A case report in SJIA and another study reporting a case of SJIA-MAS and SJIA-LD have described significant clinical improvement or even complete remission within 3 months of tofacitinib treatment [86, 87]. In another recent case report, a 4-year-old SJIA-LD patient was successfully treated with ruxolitinib, achieving significant clinical improvement that also allowed steroid tapering [88]. Two other juvenile patients with refractory dermatomyositis and associated LD responded remarkably to tofacitinib [89].

Further case reports of adult patients with dermatomyositis-associated LD describe improvement with tofacitinib treatment [90], and in line with this a single-center, open-label clinical study (Chinese Clinical Trial Registry number, ChiCTR-1800016629) evaluating the efficacy of tofacitinib in patients with dermatomyositis-LD showed improved survival and respiratory symptoms [91].

IFN crosstalk can affect SJIA therapeutics

A key consideration when using agents such as jakinibs is the significant crosstalk between Toll-like receptors (TLRs), type I and II IFNs, NFκB and STAT signalling. Type I and type II IFNs are able to prime the chromatin structure to enable robust transcriptional responses to TLR signalling, inducing sustained transcription factor binding to TNFα, IL-6 and IL12 gene loci [92–94]. Furthermore, IFN-γ can augment TLR responses (such as production of TNFα and IL-6) by IFN-γ-induced suppression of the anti-inflammatory cytokine IL-10 and STAT3 [95] (Fig. 2).

Fig. 2.

IFN crosstalk and signalling pathway modulation

(A) Type I and II IFNs modulate the immune response by epigenomic changes. IFN-γ inhibits the anti-inflammatory response of IL-10 by interfering with STAT3 signalling. IFNα/β signalling via IFNAR and DAMPs/PAMPs signalling via TLRs can work together to induce transcriptional regulation of IL-18. (B) IFNα/β downregulates IL-1α/β by upregulation of IL-10, IL1RA and the decoy IL-1R2 receptor. IL-1 suppresses IFNα/β production via Prostaglandin E2 (PGE2) upregulation. Anakinra/Canakinumab, which blocks IL-1 signalling, can thus promote increased IFNα/β levels.

Another study showed type I IFNs and NFκB have overlapping antiviral functions and NFκB is able to mediate effective ISG induction independent of type I IFNs [96]. Two previous studies have also suggested that unconventional transcription initiation complex assembly by both STAT and NFκB regulate nitric oxide synthase expression [97] and IL-18 gene expression [86] (Fig. 2). The latter study suggests that co-induction of IL-18 by IFN-α/β, and not IFN-γ-mediated STAT signalling and DAMP or PAMP mediated TLR-NFκB signalling could have significant implications for the disease pathogenesis in SJIA-MAS and SJIA-LD.

Of interest is the cross-regulation of IL-1β and type I IFNs, particularly as IL1-inhibition has now become the mainstay of treatment in SJIA patients. Type I IFNs attenuate IL-1α/β signalling through induction of anti-inflammatory IL-10, IL1RA (the natural IL1 receptor antagonist) and by regulating the decoy receptor IL-1R2. On the other side, IL-1α and IL-1β limit type I IFN production through direct transcriptional downregulation and Prostaglandin E2 production [98, 99]. Inhibition of IL-1 by canakinumab or the recombinant IL-1 receptor antagonist anakinra could thus lead to increased levels of type I IFN signalling (as previously observed in subsets of SJIA patients [37, 52]) which could in turn augment IL-18 production, setting the stage for massive IFN-γ driven hyperinflammation and MAS and LD [86] (Fig. 2). In fact, it is conceivable that in this subset of patients, dysregulation of signal transduction crosstalk contributes to the pathogenesis of hyperinflammation.

Conclusion

With jakinibs and monoclonal therapies targeting IFNs, promising therapies for SJIA and its complications SJIA-MAS and SJIA-LD may be on the horizon. Current ongoing trials in JIA and SJIA will provide more data about their efficacy and safety. The potential implications for the role of biological therapies in contributing to the development of SJIA-LD have the power to dramatically reshape the treatment landscape for SJIA and its complications. The subset of patients affected will have to be closely investigated to understand what sets them apart, and further study is required to evaluate whether they have an increased risk of developing SJIA complications. More so, there is urgent need for unravelling the origin of pathogenesis and particularly the distinct roles of type I and II IFNs in SJIA-MAS and LD to guide future decisions for novel alternative treatments.