One of the most intriguing features of systemic-onset juvenile rheumatoid arthritis (JRA) is its association with macrophage activation syndrome (MAS), a life-threatening complication caused by excessive activation and proliferation of T cells and macrophages (1–14). Such activation leads to an overwhelming systemic inflammatory reaction. The pathognomonic features of this syndrome are observed in bone marrow aspirates: numerous well-differentiated macrophages actively phagocytosing hematopoietic elements. Such cells may infiltrate almost any organ in the body and may account for many of the systemic features of this syndrome, including cytopenias and coagulopathy.
MAS is a daunting clinical challenge. In fact, it has been a major source of mortality in pediatric rheumatology. It is becoming increasingly clear, however, that the understanding of the pathophysiology of MAS not only may define better treatment and improve the outcome, but also may provide important clues to the discovery of new pathways involved in the down-regulation of cellular immune responses in humans.
MAS has strong clinical similarities with familial hemophagocytic lymphohistiocytosis (HLH). In familial HLH, the uncontrolled proliferation of T cells and macrophages has been recently associated with decreased natural killer (NK) cell and cytotoxic cell functions secondary to mutations in the gene encoding perforin (PRF1). Interestingly, similar immunologic abnormalities have also been documented in patients with MAS (15, 16). Furthermore, increasing evidence suggests that low NK cell function associated with abnormal levels of perforin expression may be a feature that distinguishes patients with systemic JRA from those with other clinical forms of JRA (17, 18). If confirmed, these observations not only will lead to an explanation for the increased incidence of MAS in systemic-onset JRA but also will provide clues for further mechanistic studies of this syndrome. The purpose of this review is to summarize the current understanding of the pathophysiology of MAS as well as its relationship to other hemophagocytic syndromes.
In the early 1980s, several reports described patients with systemic-onset JRA in whom a severe coagulopathy resembling disseminated intravascular coagulation (DIC) developed. Such coagulopathy was often associated with mental status changes, hepatosplenomegaly, increased serum levels of liver enzymes, and a sharp fall in blood cell counts and the erythrocyte sedimentation rate (ESR) (2–4). In 1985, Hadchouel et al (5) linked these symptoms to massive proliferation of activated non-neoplastic macrophagic histiocytes with prominent hemophagocytic activity. The term “macrophage activation syndrome” was eventually introduced in 1993 by Stephan et al, in a followup report originating from the same center (6). Over the following years, several more reports from various countries described a number of patients with very similar symptoms (7–12). Although MAS has also been observed in a small number of patients with polyarticular JRA (13, 14) and those with some other rheumatic diseases (14, 19, 20), it is seen most commonly in patients with the systemic form of JRA.
Because excessive activation and proliferation of tissue macrophages, or histiocytes, exhibiting hemophagocytic activity (often triggered by infections) is a pathognomonic feature of MAS, the term reactive hemophagocytic lymphohistiocytosis (HLH) has been preferred by some authors to classify this condition (14, 21, 22). This term suggests that MAS may belong to a group of histiocytic disorders collectively known as HLH. HLH is a more general term that describes a spectrum of disease processes characterized by accumulations of histologically benign well-differentiated mononuclear cells with a macrophage phenotype (23, 24). Because such macrophages represent a subset of histiocytes that are distinct from Langerhans' cells, this entity should be distinguished from Langerhans' cell histiocytosis as well as from other dendritic cell disorders.
In the contemporary classification of histiocytic disorders, HLH is further subdivided into primary, or familial, HLH and secondary, or reactive, HLH. Clinically, however, these disorders may be difficult to distinguish from each other. Familial HLH is a constellation of rare autosomal recessive immune disorders. The clinical symptoms of familial HLH usually become evident within the first 2 months of life, although initial presentation as late as 22 years of age has been reported as well (25). Secondary HLH tends to occur in older children and more often is associated with an identifiable infectious episode, most notably Epstein-Barr virus (EBV) or cytomegalovirus (CMV) infection. The group of secondary hemophagocytic disorders also includes malignancy-associated HLH.
Similar to the clinical course of MAS, the clinical course of the most typical form of HLH is characterized by persistent fever and hepatosplenomegaly. Neurologic symptoms can complicate and sometimes dominate the clinical course. Hemorrhagic rash and lymphadenopathy are observed less frequently. Laboratory findings such as cytopenias (particularly thrombocytopenia), elevated liver enzyme levels, hypertriglyceridemia, and hypofibrinogenemia also overlap with MAS. Finally, as in MAS, hemophagocytosis in bone marrow is a hallmark of HLH (Table 1). Despite all of these clinical similarities, the exact relationship between MAS and HLH is not understood.
Epidemiologic studies of MAS have been complicated by the lack of well-defined diagnostic criteria. In most cases reported in the literature, the diagnosis of MAS was initially suspected based on the development of cytopenias and coagulopathy and then was confirmed at biopsy by the demonstration of hemophagocytosis. However, many clinicians believe that MAS may simply represent a severe end of the spectrum of such abnormalities. Indeed, some degree of coagulopathy, a major clinical feature of MAS, may be present in the majority of patients with systemic-onset JRA (4, 26), a view that was reinforced by a recent report by Ravelli (27). In this review, clinical and laboratory manifestations of JRA-associated cases of MAS reported in the literature were compared with those in patients with active systemic-onset JRA. As a result, threshold values for certain measures of coagulopathy and inflammatory activity have been proposed for use as possible diagnostic criteria. The variables with the highest sensitivity and specificity were serum levels of ferritin (>10,000 mg/ml), triglycerides (>160 mg/dl), and fibrinogen (<250 mg/dl).
Despite the lack of diagnostic criteria, due to increasing awareness of MAS this syndrome is being recognized more and more frequently. According to one report originating from a tertiary level pediatric rheumatology unit, 7 of 103 patients in whom systemiconset JRA was diagnosed between 1980 and 2000 developed MAS at some point during the course of their illness (13). In another retrospective review of patients with reactive HLH, the authors noted that more than one-third of the patients fulfilled the criteria for systemic-onset JRA or adult-onset Still's disease, thus prompting them to question the distinct nature of the 2 conditions (28).
Based on our own experience and review of the literature, MAS occurs with equal frequency in boys and girls. There appears to be no racial predilection, and MAS may occur in children of almost any age. The youngest MAS patient reported to date was 12 months old (2). Although in most patients this syndrome develops sometime during the course of their primary rheumatic disease, the occurrence of MAS at the time of the initial presentation of a rheumatic illness has been described as well (2, 12). The vast majority of patients have an active primary rheumatic disease prior to the development of MAS. One report, however, described a patient whose polyarticular JRA was not active at the time of MAS presentation (7).
Although it is now evident that development of MAS can be precipitated by virtually any infectious agent (bacterial, fungal, and even parasitic) (20), viral illnesses, particularly EBV and infections caused by other members of the herpesvirus family, appear to be most commonly described (7, 13–15). In several reports, the triggering of MAS coincided with modifications in drug therapy, most notably administration of gold preparations (5, 6, 29), methotrexate (30, 31), sulfasalazine (32), and tumor necrosis factor (TNF)–blocking agents (33). These associations, however, should be interpreted cautiously, because many of the described patients had very active underlying rheumatic disease, and MAS might have been developing when therapy with the drugs was started. In most patients, triggering events are not identified.
The clinical findings in MAS are dramatic (14). Typically, patients with a chronic rheumatic condition become acutely ill with persistent fever, mental status changes, lymphadenopathy, hepatosplenomegaly, liver dysfunction, easy bruising, and mucosal bleeding. These clinical signs and symptoms are associated with a precipitous fall in the levels of at least 2 of 3 blood cell lines (leukocytes, erythrocytes, and platelets). A decline in the platelet count is usually an early finding. Because bone marrow aspiration typically reveals significant hypercellularity and normal megakaryocytes, such cytopenias do not seem to be secondary to inadequate production of cells. Increased destruction of the cells by phagocytosis and consumption at the inflammatory sites are more likely explanations.
A precipitous fall in the ESR is another characteristic laboratory feature, which probably reflects the degree of hypofibrinogenemia secondary to fibrinogen consumption (27) and liver dysfunction (6, 9). In fact, liver involvement is common in MAS. Significant hepatomegaly is frequently present. Mild jaundice develops in some patients. Liver function tests often reveal high serum transaminase activity and mildly elevated levels of serum bilirubin. Moderate hypoalbuminemia has been reported as well (2). Serum ammonia levels are typically normal or only mildly elevated.
Encephalopathy is another frequently reported clinical feature of MAS. Mental status changes, seizures, and coma are the most common manifestations of central nervous system (CNS) disease (5, 6). Cerebrospinal fluid pleocytosis with mildly elevated protein levels has been noted in some studies (5). Significant deterioration in renal function has been noted in several series (6, 13) and in one study was associated with particularly high mortality (13). Pulmonary infiltrates have been mentioned in several reports (7, 9), but the extent to which they can be attributed to MAS is not clear.
Additional laboratory findings in MAS include highly elevated serum levels of triglycerides and ferritin. The elevation of ferritin is particularly marked (>10,000 ng/ml in most patients) (27) and appears to parallel the degree of macrophage activation (34).
A hemorrhagic syndrome resembling DIC is the most striking abnormality in MAS (2–14). Hemorrhagic skin rashes (from mild petechiae to extensive ecchymotic lesions), epistaxis, hematemesis secondary to upper gastrointestinal bleeding, and rectal bleeding are the most commonly observed clinical features caused by coagulation abnormalities in MAS (2–6). Further laboratory evaluation reveals prolonged prothrombin and partial thromboplastin times, marked hypofibrinogenemia, and moderate deficiency of vitamin K–dependent clotting factors. A decrease in factor V levels is usually mild. Fibrin degradation products may be present as well.
In early reports, the coagulation abnormalities observed in this syndrome were interpreted by several authors as “consumption coagulopathy” triggered by the vasculopathic component of the disease (2). Indeed, “low level DIC,” possibly reflecting endothelial cell damage, may be a part of active systemic JRA (26). Other investigators, however, proposed that severe liver dysfunction induced by macrophages infiltrating the liver parenchyma is more relevant to the pathogenesis of coagulation abnormalities (6, 9). Indeed, liver disease may result in a complex coagulopathy caused by decreased synthesis of clotting factors, such as fibrinogen. The liver also produces inhibitors of coagulation such as antithrombin III, protein C, and protein S, and is the clearance site for activated coagulation factors and fibrinolytic enzymes. Thus, patients with severe liver dysfunction are predisposed to DIC and may develop systemic pathologic fibrinolysis. For these reasons, coagulation defects in advanced liver disease are often difficult to distinguish from those in DIC.
Prahalad et al (12) described a patient with MAS in whom severe acute hemorrhagic diathesis developed in the absence of significant liver dysfunction. The vasculopathic component in the hemorrhagic skin lesions was not prominent. The authors suggested that the coagulopathy in that patient might have been related to the procoagulant activity of the activated macrophages. Indeed, in an inflammatory response, activated macrophages can be induced to produce the hemostatic tissue factor. In turn, the tissue factor expressed in macrophages and on TNFα-stimulated endothelial cells has been shown to be central to the pathogenesis of DIC in both experimental and clinical conditions (35). A strong correlation between serum levels of soluble TNF receptors and prolongation of the partial prothrombin time in patients with systemic-onset JRA (36) suggests that increased TNFα activity observed in MAS (6, 10, 37) further contributes to the development of the coagulation abnormalities.
The most common histopathologic finding in patients with MAS is tissue infiltration with T lymphocytes and cytologically benign yet actively phagocytic macrophages (Figure 1). Although demonstration of macrophages phagocytosing hematopoietic elements in the bone marrow or lymph nodes is virtually diagnostic, negative results may be reported due to sampling difficulties or timing of the procedure (13). The macrophages may also be found in various organs and may account for many of the systemic features of this syndrome. Postmortem evaluation of one patient with MAS revealed extensive macrophagic infiltration of the heart, adrenal glands, liver, pancreas, and meninges (20). In addition to revealing sinusoidal and periportal infiltration with macrophages, histologic evaluation of the liver often reveals severe diffuse fatty changes (5). The development of fatty changes in the liver may be related to the metabolic effects of TNFα. TNFα has been shown to stimulate hepatic lipogenesis and to inhibit synthesis of lipoprotein lipase, an enzyme needed to release fatty acids from circulating lipoproteins so that they can be used by the tissues. The same mechanism also appears to be responsible for the high serum levels of triglycerides seen in MAS patients.
Two recent reports described patients with systemic-onset JRA/MAS who also had necrotizing histiocytic lymphadenopathy consistent with Kikuchi's disease (15, 38). Given the rarity of both conditions, this association may not be random. The most common cutaneous manifestations of MAS are panniculitis and purpura (12, 39). Most biopsy specimens show edema and hemorrhage associated with a mononuclear cell infiltrate and numerous macrophages occasionally showing hemophagocytosis. Schuval et al described 2 patients with histiocytic cytophagic panniculitis associated with systemic manifestations such as fever and hepatosplenomegaly (40). At some point during the course of their disease, acute pancytopenias developed in both patients and responded to cyclosporin A. Although these patients did not have apparent coagulation abnormalities, activation of macrophages did play a role in the pathogenesis of their diseases, suggesting that there may be a significant overlap between the pathogenic mechanisms involved in MAS and histiocytic cytophagic panniculitis.
Immunologic abnormalities in familial HLH
The pathogenic mechanisms involved in the development of MAS are not known. However, immunologic studies of the clinically similar entity familial HLH suggest that there is an underlying abnormality in immunoregulation that contributes to the lack of control of an exaggerated immune response (24, 41). The clinical findings during the acute phase of HLH can be explained largely as a consequence of the prolonged production of cytokines and chemokines presumably originating from activated macrophages and T cells. An excess of circulating interleukin-1β (IL-1β), TNFα, IL-6, and interferon-γ (IFNγ) is likely to contribute to the early and persistent findings of fever, hyperlipidemia, and endothelial activation responsible in part for coagulopathy, as well as later sequelae including hepatic triaditis, CNS vasculitis and demyelination, and bone marrow hyperplasia. Hemophagocytosis, the pathognomonic feature of the syndrome, is a hallmark of cytokine-driven excess activation of macrophages (24).
The most consistent immunologic abnormality reported in these patients has been impairment of cytotoxic functions. Thus, it has been demonstrated that most patients with familial HLH have normal numbers of B lymphocytes and normal serum immunoglobulin levels (42). The majority of these patients have surprisingly normal absolute lymphocyte counts and normal distribution of mature T cell subsets. In contrast, NK cell function is markedly decreased or absent in virtually all patients. Cytotoxic activity of CD8+ cells is also defective (42, 43). In ∼40% of patients with familial HLH, these abnormalities have been associated with mutations in the gene encoding perforin, a protein that mediates cytotoxic activity of NK cells and T cells (44). Patients with familial HLH usually have normal numbers of NK cells but either very low or absent perforin expression in all cytotoxic cell types, including NK cells. Therefore, the cytolytic activity of such cells is impaired.
Patients with virus-associated HLH also have very low or absent cytolytic NK cell activity. However, in contrast to familial HLH, this phenomenon appears to be related to profoundly decreased numbers of NK cells rather then impaired perforin expression. In fact, perforin expression in both CD8+ and CD56+ cytotoxic cells is often mildly increased (45). It appears that NK cell function may completely recover in some of these patients after the resolution of the acute phase of the syndrome (Kogawa K: personal communication).
Interestingly, NK cells are also affected in the Chédiak-Higashi syndrome, another disease associated with the development of lymphohistiocytic expansion and hemophagocytosis. In this disease, absolute numbers of NK cells in peripheral blood are normal, but the cells are characterized by the presence of abnormal granules containing perforin in the cytoplasm. The genetic defect in this condition appears to be a mutation in the gene encoding one of the cytoskeletal proteins, leading to impaired ability to mobilize perforin (46). As a result, the cytolytic activity of NK cells and cytotoxic CD8+ cells is greatly diminished (47).
Two other genetic conditions associated with hemophagocytic lymphohistiocytosis are Griscelli syndrome and X-linked lymphoproliferative disease. In Griscelli syndrome, hemophagocytic complications are related to the deficiency of Rab27A, a cytoplasmic protein involved in cytotoxic granule release (48), while X-linked lymphoproliferative disease is caused by a mutation in an adaptor protein, SH2DIA, that has been implicated in the regulation of T cell activation and NK cell effector function (49, 50). In both syndromes, cytotoxic functions are greatly impaired (50).
Perforin expression and NK cell function in systemic-onset JRA and MAS
Strong clinical similarities between HLH and MAS prompted several groups of investigators to examine perforin expression and NK cell function in patients with systemic-onset JRA and MAS. One recent study demonstrated reduced perforin expression in NK cells and in 2 subsets of cytotoxic CD8+ T lymphocytes (CD45RA−,CD28− and CD45RA+,CD28−) in patients with active systemic JRA compared with other forms of JRA and healthy controls (16). Interestingly, in 4 patients perforin expression returned to normal levels after autologous hematopoietic stem cell transplantation. This observation suggests that low perforin expression is likely to be induced by certain cytokines or chemokines that are abundant in active systemic JRA. Based on similarities with familial HLH, the authors also suggested that reduced perforin expression might be responsible for the increased incidence of MAS in systemic-onset JRA (16).
Another study focused on the assessment of NK cell function and perforin expression in 7 patients with MAS presenting as a complication of systemic-onset JRA (15). NK cell activity in peripheral blood samples collected during the acute stage and after the resolution of MAS was profoundly depressed in all patients. In some patients, low NK cell activity was associated with very low numbers of NK cells but mildly increased levels of perforin expression in NK T cells and cytotoxic CD8+ T lymphocytes, a pattern somewhat similar to that in virus-associated HLH. In contrast, in other patients, very low NK cell activity was associated with only mildly decreased numbers of NK cells but very low levels of perforin expression in all cytotoxic cell types, a pattern indistinguishable from that in carriers of familial HLH. Remarkably, most of the patients with low perforin expression had a history of previous episodes of MAS.
It has also been reported that decreased absolute numbers of NK cells (17) and decreased NK cell function (18) might be features that distinguish patients with systemic JRA from those with other forms of JRA. This observation may be another clue to the understanding of the reasons for the increased incidence of MAS in patients with systemic-onset JRA. These data also suggest that the pathway common to both MAS and HLH is likely to be associated with abnormal cytolytic function (see Table 2).
Table 2. Immunologic abnormalities in HLH and MAS*
Biallelic mutations in PRF1 gene
Normal PRF1 gene
NK (natural killer) cell function is defined as cytolytic activity against the K562 cell line. HLH = hemophagocytic lymphohistiocytosis; MAS = macrophage activation syndrome.
NK cell function
NK cell number
Normal or ↓
% of perforin-expressing NK cells (CD56+/CD16+)
Normal or ↓
% of perforin-expressing CD8+ cells
Normal or ↓
NK cell function and cellular immune responses
The exact mechanisms that would link deficient NK cell and cytotoxic T lymphocyte functions with expansion of activated macrophages are not clear. Two alternative explanations have been suggested in the literature. One is related to the fact that HLH/MAS patients appear to have a diminished ability to control some infections (41, 51). More specifically, NK cells and cytotoxic T lymphocytes fail to kill infected cells and, thus, fail to remove the source of antigenic stimulation. Such persistent antigen stimulation leads, in turn, to persistent antigen-driven activation and proliferation of T cells associated with escalating production of cytokines that stimulate macrophages. The fact that HLH/MAS episodes are often triggered by viruses from the herpesvirus group supports this hypothesis. The herpesviruses, including CMV and EBV, have evolved evasion mechanisms to down-regulate or sequester class I major histocompatibility complex (MHC) molecules, thus preventing efficient cytolytic CD8+ T cell responses. In contrast, such down-regulation of the expression of class I MHC molecules serves as an activating signal for NK cells that triggers their cytolytic activity against infected cells. Therefore, NK cells become particularly important in the defense against herpesviruses. Consistent with this idea, infection of NK-depleted or perforin-deficient mice with murine CMV does result in an exaggerated immune response associated with more persistent expansion of cytotoxic CD8+ T cells that secrete IFNγ, an important macrophage activator (52, 53).
In many cases of MAS, however, attempts to identify an infectious trigger have not been successful, and some episodes appear to be triggered by modifications in drug therapy rather than infection. Furthermore, the importance of NK cells and perforin-based systems in the down-regulation of the cellular immune responses has been demonstrated in experimental animal systems, in which immune responses were elicited by anti-CD3 antibodies or staphylococcal toxins instead of viruses. For instance, Kagi et al demonstrated that the injection of staphylococcal enterotoxin B into perforin-deficient mice resulted in dramatically increased selective expansion and prolonged persistence of CD8+ but not CD4+ staphylococcal enterotoxin B–reactive T cells (54). The results of these experiments suggest that there likely is a more direct effect of perforin-based systems on the survival of activated lymphocytes.
It has been hypothesized by some authors that abnormal cytotoxic cells may fail to provide appropriate apoptotic signals for removal of the antigen-presenting cells and/or activated T cells after infection is cleared (55). Such T cells may continue to secrete cytokines, including IFNγ and granulocyte–macrophage colony-stimulating factor (GM-CSF), 2 important macrophage activators. Subsequently, the sustained macrophage activation results in tissue infiltration and in the production of high levels of TNFα, IL-1, and IL-6, which play a major role in the various clinical symptoms and tissue damage. A report on successful treatment of JRA-associated MAS by cyclosporin A with transient exacerbation of MAS by conventional-dose G-CSF supports this hypothesis (11).
MAS is a serious complication of systemic JRA and is associated with considerable morbidity and mortality. Therefore, early recognition of this syndrome and immediate therapeutic intervention to produce a rapid clinical response are critical. Thus, in a patient with persistently active systemic-onset JRA, a fall in the ESR and platelet count, particularly in combination with an increase in D-dimer levels, should raise suspicion of impending MAS. Prompt administration of more aggressive treatment in these patients may, in fact, prevent development of the full-blown syndrome. In our clinic, to achieve rapid reversal of coagulation abnormalities, we often start with intravenous methylprednisolone pulse therapy (30 mg/kg/day for 3 consecutive days) followed by 2–3 mg/kg/day in 4 divided doses. After normalization of hematologic abnormalities and resolution of coagulopathy, we taper steroids slowly to avoid relapses of macrophage activation. Not uncommonly, however, MAS appears to be corticosteroid resistant, with deaths being reported even among patients treated with massive doses of steroids.
Parenteral administration of cyclosporin A in patients with corticosteroid-resistant MAS was recently proposed by Mouy et al (9). From the primary effect of cyclosporin A, largely but not entirely confined to T cells, a wide variety of other effects are mediated, leading to profound and therapeutically useful immunosuppression (56). In the patients with MAS described by Mouy et al (9) and Ravelli et al (10), parenteral administration of cyclosporin A (2–8 mg/kg/day) not only provided rapid control of symptoms but also allowed avoidance of excessive use of corticosteroids.
There have been anecdotal reports on the use of etoposide, a podophyllotoxin derivative that inhibits DNA synthesis by forming a complex with topoisomerase II and DNA. For instance, etoposide was successfully used to treat EBV-induced MAS in a 21-year-old woman with longstanding systemic-onset JRA (8) but induced severe bone marrow suppression in another patient with MAS (6).
Of note is the fact that the combination of steroids, cyclosporin A, and etoposide is the major component of the HLH treatment protocol (HLH-94) developed by the international Histiocyte Society (57). This protocol includes a combination of etoposide and CNS-penetrating dexamethasone (with or without intrathecal methotrexate), followed by a maintenance course of cyclosporin A and less frequent pulses of etoposide once clinical remission has been established. In accordance with the protocol, patients with familial HLH and patients who experience a relapse after initially responding to HLH-94 should proceed to definitive therapy with allogeneic hematopoietic stem cell transplantation.
A suggested role for TNFα in the development of coagulation abnormalities provided a rationale for the use of TNF inhibitors in corticosteroid-resistant cases of MAS (12, 58), which resulted in a good clinical response. However, other reports describe patients in whom MAS developed while they were being treated with TNF inhibitors (15, 33). In most of those cases, MAS was triggered by infections. Because TNF-inhibiting agents may increase susceptibility to infections, including those that trigger MAS, these medications should be considered only when an infectious etiology has been ruled out. In other words, it appears that although TNF inhibitors may help control some of the downstream pathways of MAS, these drugs do not provide protection against triggering of this syndrome.
The analysis of pathophysiologic similarities between HLH and MAS leads to the hypothesis that impaired cytotoxic functions and the lack of an immunoregulatory role of NK cells are relevant to the development of both syndromes. It appears likely that an intrinsic cytolytic defect is responsible for the failure to down-regulate cellular immune responses triggered by either infection or some other environmental factors. Furthermore, profound NK dysfunction may be a feature that distinguishes systemic JRA from other clinical forms of JRA as well as other rheumatic conditions. If this hypothesis were proven, it would explain the increased incidence of MAS in patients with systemic-onset JRA and provide a basis for further mechanistic studies of MAS and HLH. Both conditions are rare; therefore, cooperative efforts from pediatric rheumatologists in the US and around the world will be necessary to better understand the nature of MAS and its relationship to other HLH syndromes.