New classification criteria for juvenile idiopathic arthritis

Juvenile idiopathic arthritis (JIA) refers to a group of heterogeneous diseases sharing unknown original chronic inflammatory arthritis over 6 weeks. The classification system proposed by the International League Against Rheumatism (ILAR) is accepted worldwhile.1 The unmet needs of JIA include a more refined classification system in line with the adult forms of arthritis, therapies specific for individual JIA categories based on their pathogenesis, and prevention of the long-term consequences.2 The main criticisms of the ILAR JIA classification are as follows. The number of joints involved and the presence of psoriasis may not be useful to define homogeneous entities in children.3 The number of joints involved may simply represent disease severity but may not be a reliable major criterion in the categories due to experts' subjectivity to evaluation results. The different categories include heterogeneous conditions. The homogeneous disease in children, which is not observed in adults, could not be identified by the current classification. Exclusion criteria increase the number of undifferentiated arthritis patients. Occasionally, arthritis can occur several years after the appearance of systemic symptoms in systemic JIA (sJIA). Antinuclear antibody (ANA) positivity, presence of iridocyclitis, asymmetry of arthritis, and onset age may be more meaningful for defining a homogeneous disease entity.4 Some experts claim that diagnosis should not be ruled out if patients have the same systemic characteristics consistent with sJIA, although they never develop arthritis. Based on this statement, the newly proposed JIA classification states that sJIA is equivalent to adult-onset Still disease (AOSD). Therefore, the presence of arthritis is no longer mandatory. In addition to fever, the presence of classic rash plus at least two secondary criteria is sufficient for diagnosis.5 The treatment goal of JIA is to achieve continuous remission or low disease activity. Disease course and JIA subtype are two factors affecting the remission rate. Patients with sJIA had the highest frequency of remission off medications, whereas rheumatoid factor (RF) -positive polyarthritis patients had the lowest.6 As JIA classification changes in about one-third of children during the disease course, JIA classification should be revised during follow up.7 About 50% of patients could not effectively be transferred to an adult rheumatologist so they were at risk of poor outcomes.8 The ILAR classification does not distinguish the chronic arthritis observed in both adults and children from those forms of the disease that may be unique to childhood. This distinction, underlying obvious pathogenetic and therapeutic implications, is very important, not only for the transition process from childhood to adult care but also to new drug registration. If RF-positive JIA is regarded as the same disease as adult RF-positive rheumatoid arthritis, the optional therapeutic plan is shared. Controlled clinical trials testing new drugs could be first performed in adults, then convenient open-label studies to assess proper dose and long-term safety in children could be performed after safety and efficacy have been demonstrated.9, 10 In this way, the procedure to obtain new drugs registration in children could be greatly accelerated.11 The ILAR classification criteria focus on the importance of extra-axial manifestations such as peripheral arthritis and enthesitis. In contrast, classification of spondyloarthritis (SpA) in adults focuses on the presence of axial disease and peripheral disease. Due to their different classification criteria, the US Food and Drug Administration (FDA) had waived the studies for new medications for “axial spondyloarthropathies including ankylosing spondylitis” in children until July 2020. Although current JIA treatment guidelines recommend the use of biologic disease-modifying anti-rheumatic drugs as part of the early treatment for patients with enthesitis-related arthritis (ERA), none of the FDA-approved therapies for peripheral SpA or non-radiographic axial SpA have been studied or are labeled for use in children with ERA.12 A large-scale genomic study of sJIA demonstrated that sJIA did not share heritable risk factors with oligoarticular and polyarticular forms of JIA. The subtype-specific risk factors (single nucleotide polymorphisms, human leukocyte antigen [HLA] alleles and HLA haplotypes) are not shared between different subtypes.13 However, characteristics of JIA patients fit the classification criteria of adults in transition period. Oligoarticular JIA and RF-negative polyarticular JIA may be pediatric entities, whereas the other types of JIA tend to meet the respective adult classification criteria.14 The Yamaguchi criteria may be useful for classifying childhood sJIA, particularly in the “pre-arthritic”, pure, systemic disease phase.15 Most patients with ERA had active disease even in adulthood, and nearly all ERA patients fulfilled adult SpA criteria and the stricter modified New York criteria for ankylosing spondylitis were fulfilled by 75% of patients.16 Pediatric patients currently lack any taxonomic definition, although they are are classified as AOSD using Yamaguchi criteria with minor modifications.3 Psoriatic arthritis (PsA) at a young age is too similar to oligoarthritis to distinguish between them, so removal of early-onset PsA from the classification was proposed.17 The category of RF-negative polyarticular JIA contains at least two different disease subgroups. The ANA-positive subgroup has characteristics similar to oligoarticular-onset JIA, whereas the ANA-negative subgroup exhibits features similar to patients with seronegative adult polyarthritis. ANA-positive patients with persistent oligoarticular, extended oligoarticular, or polyarticular JIA have the same disease entity and differ only in disease severity.4 Corroborating studies suggest that HLA associations and peripheral blood transcriptomic signatures differ at the boundary of 6 years old. These patients are negative for RF and anti-citrullinated protein antibodies but are typically positive for ANAs. This population is also at highest risk of chronic anterior uveitis,18 which can occur in young children without arthritis, raising the possibility that this hallmark feature reflects the pediatric substrate rather than a unique arthritis-associated biology.19 A new disorder has been suggested by the proposed criteria, called early-onset ANA-positive JIA, which appears to represent a condition unique to the pediatric population and includes up to 50% of all JIA cases.3 A new classification should be able to distinguish arthritis cases that are typical of children from those that observed in both childhood and adults.20 After three Delphi Web rounds of interactions and the Nominal Group Technique consensus conference, a new evidence-based classification model is proposed by the Pediatric Rheumatology International Trials Organization (PRINTO; Table 1). The first step reflects the effort to identify, based on available clinical and laboratory measures worldwide, distinct and homogeneous disorders characterized by chronic childhood arthritis, and to facilitate the validation of the proposed criteria. The identified disorders are considered the juvenile counterparts of diseases that are also observed in adults (systemic JIA, RF-positive JIA, enthesitis/spondylitis-related JIA) or are typical only of children (early-onset ANA-positive JIA), the disorders provisionally classified as “others” during the first step will be better characterized during the prospective data collection from at least 1000 new-onset patients.3 The main difference between the PRINTO and ILAR classification criteria are summarized in Table 2. Fever (at least 3 consecutive days, reoccurring over 2 weeks) and One of major criteria: (1) evanescent (nonfixed) erythematous rash; (2) arthritis, or Two of minor criteria: (1) generalized lymph node enlargement and/or hepatomegaly and/or splenomegaly; (2) serositis; (3) arthralgia lasting 2 weeks or longer (in the absence of arthritis); (4) leukocytosis (≥15 000/mm3) with neutrophilia. Arthritis for ≥6 weeks, and RF (two positive tests, at least 3 months apart) or CCP (one positive). Peripheral arthritis and enthesitis, or Arthritis or enthesitis plus two of the following: (1) sacroiliac joint tenderness; (2) inflammatory back pain; (3) presence of HLA-B27 antigen; (4) acute (symptomatic) anterior uveitis; and (5) history of SpA in a first-degree relative. Arthritis for ≥6 weeks; Early-onset (≤6 years); Presence of ANA 2 positive, titer ≥1/160 (tested by immunofluorescence, at least 3 months apart). Exclusions are systemic JIA, RF-positive arthritis, and enthesitis/spondylitis-related JIA. Arthritis for ≥6 weeks Exclusions are systemic JIA, RF-positive arthritis, enthesitis/spondylitis-related JIA and Early-onset ANA-positive JIA. 7 (Systemic arthritis; Enthesitis-related arthritis; Rheumatoid factor-positive polyarthritis; Oligoarthritis; RF-negative polyarthritis; Psoriatic arthritis; Undifferentiated arthritis.) 5 (Systemic JIA; RF-positive JIA; Enthesitis/spondylitis-related JIA; Early-onset ANA-positive JIA; Other JIA) Familial aggregation of JIA revealed genetic influence in the pathogenesis of JIA. Psoriasis, autoimmune thyroid disease, rheumatoid arthritis, and ankylosing spondylitis were the most often reported autoimmune diseases in parents of patients with JIA.21 One father had PsA and his son had AOSD; they were both HLA-B39-positive, which likely played an important pathogenic role.22 Similarity between HLA and non-HLA is strong evidence of the consistency between the types of arthritis, while the dissimilarity is strong evidence of the opposite, making genetics the touchstone for correct identification of subsets.18 Early studies revealed that siblings or twins of JIA patients have an increased risk up to 15%–30%. The onset age and disease duration among JIA twins are highly consistent, suggesting that JIA is a polygenic genetic disease. Numerous susceptibility loci have been identified through genome-wide association studies of autoimmune diseases. With the application of molecular genetic and genomic technologies in JIA research, some disease-causing genetic variants have been identified in JIA families. The functional genes in these loci have significantly increased the knowledge about the pathogenesis underlying this complex disease.23 Many rheumatic conditions may be syndromes rather than diseases. Some monogenic diseases can perfectly mimic the clinical features of a rheumatic disease.2 Single nucleotide variants are statistical associations that tag the region and are often located in noncoding regions of the genome. Consequently, pinpointing and characterizing the causal variant remains a major challenge. Large sequencing studies have not been performed in JIA.2 Systemic JIA bears a unique genetic architecture, indicating that its underlying pathophysiological mechanisms are significantly divergent from other forms of JIA. It is important to convert from broad classifications based on non-specific clinical observations to molecular and genetic data in establishing diagnoses, as well as pathophysiology for personalized medicine.13 Janus kinase inhibitor may be a potential therapeutic option for some patients with refractory systemic-onset JIA and AOSD.24 Tofacitinib, a novel Janus kinase inhibitor, showed sustained clinical improvement in refractory polyarticular course JIA patients.25 However, given the limited number of patients, it is difficult to confirm the efficacy of Janus kinase inhibitors in sJIA patients. Success of Janus kinase inhibitors in AOSD may translate to sJIA given their similarities.26 The new evidence-based classification needs further validation in new-onset JIA patients, the final nomenclature will be discussed by a team of specialists. Genetic analysis, the therapeutic response to cytokine blockers, omic-based stratification, and machine-learning techniques for data analysis can improve classification, and deserve further research. All authors have read and approved the final manuscript. Kexin Chen, Haisheng Zeng and Galymzhan Togizbayev made the equal contribution to the article. A special acknowledgement should be shown to Professor Alberto Martini from Pediatric Rheumatology International Trials Organization, for his wonderful lecture in 2022 Sino-A PAMBI Pediatric Rheumatology and Immunology Summit Conversation. The authors declare there is no conflict of interest.