Several classification systems for chronic arthritis in childhood have been proposed over the past few decades [5]. The older terms “juvenile rheumatoid arthritis,” commonly used by the American College of Rheumatology (ACR), and “juvenile chronic arthritis” by the European League Against Rheumatism, were replaced by the term “juvenile idiopathic arthritis” (JIA) at the Pediatric Standing Committee of the International League of Associations for Rheumatology (ILAR) in 1993 [6,7]. In the current classification system established by ILAR, JIA is further categorized into 7 distinct subtypes, based on their presentation within the first 6 months (Table 2). Most pediatric rheumatologists agree that the ILAR criteria are require revision for several reasons: the criteria have been established for more than 20 years, and the number of affected joints or the presence of psoriasis does not reflect a homogeneous immune pathogenesis [8,9]. Although there is controversy regarding the method employed to change the criteria, a consensus conference led by the Pediatric Rheumatology International Trials Organization (PRINTO) recently proposed a new classification system that is in the process of validation (Table 3) [10].

Biomarkers that are useful to classify the subtypes of JIA include rheumatoid factor (RF), human leukocyte antigen (HLA)-B27, antinuclear antibody (ANA), and anti-cyclic citrullinated peptide (CCP) antibodies.

RFs are antibodies directed against the Fc fragment of immunoglobulin G (IgG). They are heterogeneous and usually composed of IgM; therefore, most assays detect only IgM [11]. RFs cannot be used to diagnose JIA because IgM RFs are rarely found in children younger than 7 years of age, and can be detected not only in children with other autoimmune diseases but also in healthy children. Nevertheless, RFs are an important component of both the ILAR classification criteria and the newly proposed PRINTO criteria (Tables 2 and 3). RFs are most commonly detected in children with a later age at JIA onset and in those with polyarticular involvement [12]. In the PRINTO classification, RF-positive JIA is considered to be the pediatric counterpart of RF-positive rheumatoid arthritis (RA) in adults [10].

HLA-B27 is a class I surface antigen encoded by the B locus in the major histocompatibility complex, and has an affinity for antigenic peptides and presents them to cytotoxic T-lymphocytes [13]. The presence of HLA-B27 is an important part of the inclusion criteria used in the definition of enthesitis-related arthritis (ERA) based on the ILRA criteria or according to the definition of the enthesitis/spondylitis-related JIA provided by PRINTO (Tables 2 and 3). Approximately 60%∼80% of children with ERA represent positive for HLA-B27 [12]. Although HLA-B27 is frequently present in these children, the presence or absence of HAL-B27 cannot be relied upon to confirm or exclude the diagnosis of JIA. The allele is an indicator of risk rather than a diagnostic test. HLA-B27 also has a strong association with ankylosing spondylitis (AS), which is notable considering that ERA is thought to be the pediatric counterpart of adult AS.

ANAs are antibodies that can be directed against cellular antigens and are detected predominantly in the nucleus and in other cellular components. ANAs may be found in 5%∼18% of healthy children who do not show evidence of arthritis [14], which limits their use for diagnostic testing. However, ANAs are most frequently found in 65%∼85% of young girls with oligoarthritis and uveitis [15,16]. These findings suggest that ANA positivity may represent a unique condition in JIA. Therefore, the PRINTO proposed a new diagnostic criterion called early-onset ANA-positive JIA that was not addressed in the ILAR criteria (Table 3) [10]. This subtype exists only in JIA and does not have a matching form in adults.

Anti-CCP antibodies are a group of autoantibodies that are directed against peptides and proteins that are citrullinated. Although the major clinical relevance of anti-CCP antibodies is their association with adult RA, they are also detected in children with JIA. In a study by Tebo et al. [17], anti-CCP antibodies were found in 73% of children with RF-positive polyarticular JIA, 19% with extended oligoarticular JIA, 13% with systemic JIA, 8% with RF-negative polyarticular JIA, 4% with ERA or persistent oligoarticular JIA, and in 2% of controls. Although anti-CCP antibodies have not yet been included in the ILAR classification criteria, they are included in the new PRINTO criteria, in which RF-positive JIA was defined as arthritis that persisted over 6 weeks, along with 2 positive tests for RF at least 3 months apart, or at least 1 positive test for anti-CCP antibodies (Table 3).

Among the JIA subtypes, systemic JIA is considered to be an autoinflammatory condition that is both clinically and immunologically distinct from the other categories of JIA and therefore requires different therapy. In addition, children with this subtype are at risk of a severe, life-threatening complication called macrophage activation syndrome (MAS). In this unique type of JIA, the concentrations of S100 proteins such as S100A8–S100A9, S100A12, and interleukin (IL)-18 are increased in the plasma of children with systemic JIA [18]; therefore, these proteins may be supportive candidate biomarkers for the diagnosis of systemic JIA.

The biomarkers most commonly used to assess disease activity in JIA are C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR). CRP is an acute phase protein of hepatic origin the expression of which increases following IL-6 secretion by macrophages and T cells. CRP levels begin to increase within hours of the onset of inflammation, with levels peaking between 1 and 3 days [14]. The plasma half-life of CRP is 19 hours, making it a more sensitive and accurate reflection of the acute phase response than ESR [19]. In addition, CRP levels normalize more rapidly than ESR in response to therapy. However, CRP does not have sufficient specificity to differentiate infectious from noninfectious inflammation in JIA. ESR is the rate at which red blood cells settle though plasma to form a sediment in whole blood in a standardized tube over a period of 1 hour. ESR is also a non-specific measure of inflammation that is included in both the Juvenile Arthritis Disease Activity Score (JADAS) and the core response variables identified by the ACR, which are the standardized and certified assessment tools of JIA disease activity. The JADAS includes 4 measures: physician global assessment of disease activity, parent/patient global assessment of well-being, active joint count, and ESR [20]. The ACR pediatric core criteria for JIA comprises 6 items: physician global assessment of disease activity, parent/patient global assessment of overall well-being, functional ability, number of joints with active arthritis, number of joints with limited range of motion, and ESR [21].

Other candidate biomarkers that are useful to measure disease activity are the S100 proteins and matrix metalloproteinase 3 (MMP3). The S100 proteins are a class of pro-inflammatory calcium-binding proteins that are involved in the regulation of calcium homeostasis, enzyme activities, energy metabolism, cell growth and differentiation, and cytoskeletal dynamics [22]. In response to injurious or infectious stimuli, they also act as damage-associated molecular pattern molecules, with increased expression and release at sites of inflammation [23]. The most studied S100 proteins in JIA are S100A8 (also known as myeloid-related protein [MRP] 8 and calgranulin A), S100A9 (also known as MRP 14 and calgranulin B), and S100A12 (also known as MRP 6 and calgranulin C). S100A8 and S100A9 often exist as a complex of S100A8 and S100A9, (also known as calprotectin, and S100A8–S100A9), which is the most prevalent form under most physiological conditions. Increased levels of S100A8–S100A9 and S100A12 in the serum and synovial fluid were shown to correlate with disease activity in JIA, suggesting that these proteins may be promising markers for monitoring disease activity [24-26].

MMP-3 is an enzyme that in humans is involved in the breakdown of extracellular matrix, and may induce degradation of cartilage and bone in arthritis [27]. Concentrations of MMP-3 in the serum were higher in children with ERA than in healthy subjects and were correlated with clinical parameters of disease activity [28,29]. Therefore, MMP-3 is expected to serve as a useful biomarker for monitoring disease activity in JIA.

RFs and anti-CCP antibodies are not only components of the classification criteria for JIA, but are also important predictors of disease prognosis. Older children with RF tend to have a longer disease duration, subcutaneous rheumatoid nodules or articular erosions, or a poorer functional class, which mimics adult RA [12,18]. In most studies, anti-CCP antibodies have been detected predominantly in patients with RF-positive polyarticular JIA [17]. The prognostic value of anti-CCP antibodies in JIA is yet to be studied, but in small studies they seem to be associated with a worse prognosis [30,31].

MAS is an aggressive and life-threatening complication of systemic inflammatory disorders. MAS occurs most frequently with systemic JIA [32]. The hallmark clinical features of MAS include high fever, hepatosplenomegaly, lymphadenopathy, hemorrhagic manifestations, and encephalopathy [33]. The characteristic laboratory features are pancytopenia, liver dysfunction, disseminated intravascular coagulation, hypofibrinogenemia, hypoalbuminemia, hyperferritinemia, hypertriglyceridemia, increased levels of lactate dehydrogenase, soluble CD25, soluble CD163, and decreased ESR levels [33]. With the recognition that MAS is clinically similar to hemophagocytic lymphohistiocytosis (HLH), some have applied the HLH-2004 diagnostic guidelines to the diagnosis of MAS (Table 4) [34]. However, the main shortcoming of using HLH-2004 guidelines is that some individual criteria may not be met for systemic JIA patients. Due to the remarkable inflammatory properties of systemic JIA, a relative decrease in leukocyte count, platelet count or fibrinogen level may be more useful for early diagnosis of systemic JIA than the absolute reduction required by the HLH-2004 standard. Therefore, the classification criteria of MAS complicating systemic JIA were developed in 2014, based on a combination of expert consensus, evidence from the medical literature, and analysis of patient data (Table 4) [34].

The S100 proteins are not only biomarkers for disease activity monitoring in JIA, but also serve as useful biomarkers to predict the response to treatment or disease flares in JIA. High serum levels of S100A8–S100A9 proteins before anti-rheumatic therapies were associated with good responses to methotrexate [35-37], and anti-tumor necrosis factor (TNF) agents [36]. In contrast, the higher levels of S100A8–S100A9 in the serum during remission were associated with a risk of relapse after discontinuation of methotrexate, which supported the hypothesis that the levels of these proteins in the serum could be useful parameters to inform the decision to discontinue treatment [37].

Similar to the findings for S100A8–S100A9, JIA patients with increased levels of serum S100A12 were also more likely to respond favorably to methotrexate or anti-TNF treatment. Increased levels of serum S100A12 during remission were associated with an increased risk of disease flare [38-40]. Similar to non-systemic JIA, increased serum levels of S100A8–S100A9 during clinical remission was also a useful biomarker to predict relapse after therapy withdrawal in systemic JIA [25,35]. IL-18 and S100A8–S100A9 serum levels may be candidate biomarkers for response to therapy in systemic JIA [41]. IL-18 serum levels increased in 4 out of 8 children who experienced failure in tapering of therapy and slightly increased in only 1 out of 7 patients who experienced successful tapering of therapy [42]. S100 proteins and IL-18 are considered for useful markers for predicting treatment response and relapse in systemic JIA and for the diagnosis of the JIA subtype.