Juvenile Idiopathic Arthritis

Arthritis is one of the most common chronic diseases of children and youth, and an important cause of short- and long-term disability. There are many causes of chronic arthritis, but the most common are those grouped under the name juvenile idiopathic arthritis . Specific information about each category of chronic arthritis is provided in individual chapters. The purpose of this chapter is to summarize general information that relates to all categories of JIA.

Classifications of Chronic Childhood Arthritis

Chronic arthritis in childhood is a complex area of study not least because of inconsistencies of classification. Since its introduction in 1995, the term juvenile idiopathic arthritis (JIA) has largely supplanted the terms juvenile chronic arthritis (JCA) and juvenile rheumatoid arthritis (JRA). However, it is necessary to understand the earlier classifications to interpret the older literature on the subject.

In the 1970s, two sets of criteria were proposed to classify chronic arthritis in childhood: those for JRA, developed and later validated by a committee of the American College of Rheumatology (ACR), and those for JCA, published by the European League Against Rheumatism (EULAR). Inconsistencies between these two classifications led to confusion, and a classification proposed and revised by the Task Force for Classification Criteria of the Pediatric Standing Committee of the International League of Associations for Rheumatology (ILAR) sought to provide an internationally agreed system of definitions to further the study of childhood arthritis. These three consensus-based classifications are compared in Table 15-1 .

TABLE 15-1

Comparison of EULAR, ACR, and ILAR Criteria for Classification of Chronic Arthritis of Childhood

Onset types 3 6 6
Course subtypes 9 None 1
Age at onset of arthritis <16 yr <16 yr <16 yr
Duration of arthritis 6 wk 3 mo 6 wk
Includes JAS No Yes Yes
Includes JPsA No Yes Yes
Includes inflammatory bowel disease No Yes No
Other diseases excluded Yes Yes Yes

ACR Criteria for Classification of Juvenile Rheumatoid Arthritis

The ACR criteria ( Box 15-1 ) defined an age limit in children, the duration of disease necessary for a diagnosis, and the characteristics of the arthritis and extraarticular disease. The requirement that age at onset of arthritis be less than 16 years is based more on practice patterns than on age-related biological variation in disease. Furthermore, although persistent objective arthritis in one or more joints for 6 weeks is sufficient for diagnosis, a disease duration of at least 6 months is required before the onset type can be confirmed (unless characteristic systemic features are present). When classification depends on the number of inflamed joints, each joint is counted separately, except for the joints of the cervical spine, carpus, and tarsus; each of these structures is counted as one joint. An affected joint is defined as one with limitation of range of motion with evidence of past or current inflammation (warmth, swelling, pain on motion, or tenderness). Joint swelling or effusion is sufficient to define an actively inflamed joint. The criteria for affected joints are based on physical examination, not imaging.

Box 15-1

Criteria for the Classification of Juvenile Rheumatoid Arthritis

  • 1.

    Age at onset <16 years

  • 2.

    Arthritis (swelling or effusion, or presence of two or more of the following signs: limitation of range of motion, tenderness or pain on motion, and increased heat) in one or more joints

  • 3.

    Duration of disease: 6 weeks or longer

  • 4.

    Onset type defined by type of disease in first 6 months:

    • a.

      Polyarthritis: ≥5 inflamed joints

    • b.

      Oligoarthritis (pauciarticular disease): <5 inflamed joints

    • c.

      Systemic onset: arthritis with characteristic fever

  • 5.

    Exclusion of other forms of juvenile arthritis

Modified from J.T. Cassidy, J.E. Levinson, J.C. Bass, et al., A study of classification criteria for a diagnosis of juvenile rheumatoid arthritis, Arthritis Rheum. 29 (1986) 274–281.

EULAR Criteria for the Classification of Juvenile Chronic Arthritis

The term juvenile chronic arthritis was proposed by EULAR in 1977 for the heterogeneous group of disorders that present as chronic arthritis in childhood of unknown cause ( Box 15-2 ). These criteria differed from the ACR criteria in three ways: (1) arthritis must have been present for at least 3 months (instead of 6 weeks); (2) juvenile ankylosing spondylitis (JAS), psoriatic arthropathy, and arthropathies associated with inflammatory bowel disease are included as separate categories; (3) the term juvenile rheumatoid arthritis was applied only to children with arthritis and rheumatoid factor (RF) positivity.

Box 15-2

Criteria for a Diagnosis of Juvenile Chronic Arthritis

  • 1.

    Age at onset <16 years

  • 2.

    Arthritis in one or more joints

  • 3.

    Duration of disease 3 months or longer

  • 4.

    Type defined by characteristics at onset:

    • a.

      Pauciarticular: <5 joints

    • b.

      Polyarticular: >4 joints, rheumatoid factor negative

    • c.

      Systemic: arthritis with characteristic fever

    • d.

      Juvenile rheumatoid arthritis: >4 joints, rheumatoid factor positive

    • e.

      Juvenile ankylosing spondylitis

    • f.

      Juvenile psoriatic arthritis

From EULAR Bulletin 4, Nomenclature and classification of arthritis in children, National Zeitung AG, Basel, 1977.

ILAR Criteria for the Classification of Juvenile Idiopathic Arthritis

In 1995, the Classification Taskforce of the Pediatric Standing Committee of ILAR proposed a classification of the idiopathic arthritides of childhood ( Box 15-3 ). This classification and its subsequent revisions were developed by consensus with the aim of achieving homogeneity within disease categories to better facilitate clinical and basic research, and to eliminate inconsistencies resulting from the use of the ACR and EULAR classifications. Like its predecessors, this classification system applies to children under the age of 16 years and is based on disease expression during the first 6 months of disease. It differs, however, in the application of exclusion criteria that eliminate overlap and improve homogeneity within the six subtype categories. The designation “undifferentiated arthritis” includes conditions that either do not meet criteria for any other category or meet criteria for more than one category. These criteria have been virtually universally adopted and are the subject of a number of studies. They were intended to be modified on the basis of emerging evidence regarding pathogenesis and disease course. Differences in nomenclature require that care be taken in interpreting the literature because the terms JRA, JCA, and JIA are often incorrectly used as if they were synonymous.

Box 15-3

Proposed Classification Criteria for Juvenile Idiopathic Arthritis

Durban, 1997

  • 1.


  • 2.


    • a.


    • b.


  • 3.

    Polyarthritis (rheumatoid factor negative)

  • 4.

    Polyarthritis (rheumatoid factor positive)

  • 5.

    Psoriatic arthritis

  • 6.

    Enthesitis-related arthritis

  • 7.

    Undifferentiated arthritis

    • a.

      Fits no other category

    • b.

      Fits more than one category

From R.E. Petty, T.R. Southwood, J. Baum, et al., Revision of the proposed classification criteria for juvenile idiopathic arthritis: Durban, 1997, J. Rheumatol. 25 (1998) 199–1994.

Criteria for Classification of Spondyloarthropathies

Wright and Moll introduced the concept of the seronegative spondyloarthritis, a grouping of a number of chronic arthritides on the basis of seronegativity for RF, absence of rheumatoid nodules, a tendency to family history of similar diseases, and the frequent involvement of the sacroiliac joints and lumbosacral spine. The later discovery of the association of the human leukocyte antigen HLA-B27 with these diseases further strengthened the legitimacy of the association. Included in the group were ankylosing spondylitis (AS), psoriatic arthritis, reactive arthritis, inflammatory bowel disease, JCA, Whipple disease, Behçet syndrome, and acute anterior uveitis. Currently, the spondyloarthritis concept includes AS, psoriatic arthritis, reactive arthritis, and the arthritis of inflammatory bowel disease. Ankylosing spondylitis is the prototype of the spondyloarthritides, and the other members are quite heterogeneous; some have a strong similarity to AS, others do not, especially in childhood. Criteria for the classification of AS require the presence of radiographic evidence of sacroiliitis as well as a history of pain at the lumbosacral junction or lumbar spine, restriction in range of motion of the spine and limitation of chest expansion. These criteria have limited utility in diagnosing or classifying early disease, or in diagnosing children with AS. Many patients who have some characteristics of AS but do not meet the diagnostic criteria for AS are included in the spondyloarthropathies as defined by Amor et al. or the European Spondylitis Study Group (ESSG). The ESSG criteria have been studied in children. Children who fall within the category of spondyloarthropathy include those who, by the ILAR criteria, have enthesitis-related arthritis (ERA), or juvenile psoriatic arthritis (JPsA). (See discussions in Chapters 19 and 20 ). The ESSG criteria also include children with arthritis associated with inflammatory bowel disease, a category not included in the ILAR criteria. There is no question that children with these diseases sometimes have or will develop sacroiliitis, and sometimes they are HLA-B27 positive, but the majority do not have these characteristics and differ significantly from each other clinically. For these reasons the ILAR criteria categorize each separately. The reader is referred to the thoughtful commentary by Colbert.

A Critique of the ILAR Criteria

The ILAR criteria have fulfilled their purpose of providing a common language for investigation and communication concerning chronic childhood arthritis of unknown cause. Some of the parameters of the ILAR criteria are entirely arbitrary in order to retain some continuity with the ACR and EULAR criteria: the age at onset is limited to children under the age of 16; the disease duration (6 weeks), the duration of the onset period (6 months, consistent with the ACR criteria); the number of inflamed joints (less than 5 or greater than 4) that defines oligoarticular and polyarticular JIA, respectively. Inflamed joints are defined clinically, whereas many studies have demonstrated that ultrasound or magnetic resonance imaging are superior to clinical examination for the documentation of joint inflammation.

Systemic JIA. The question arises as to whether systemic JIA should be included in the category of JIA, or whether it is actually an autoinflammatory disease ( Chapters 16 and 47 ). Systemic JIA is probably the same disease as Adult-onset Still’s Disease for which the Yamaguchi criteria were developed ( Box 15-4 ). These criteria have been applied to the diagnosis of children who do not meet the ILAR criteria for systemic onset JIA because they do not have arthritis at disease onset. In a study by Kumar and colleagues, of 34 children who eventually fulfilled ILAR criteria for a diagnosis of systemic JIA, 13 did not have arthritis at onset of disease, but developed objective joint disease from 15 days to 1 year later (median 30 days, mean 69.3 days ± 104.6). The ILAR classification is based on the disease characteristics in the first 6 months of disease, and many of the children who did not have frank arthritis when first evaluated did so within 6 months of fever onset, and thus would have fulfilled the ILAR criteria.

Box 15-4

The Yamaguchi Criteria for Diagnosis of Adult Onset Still’s Disease

  • Major Criteria

    • Fever of 39°C of higher, lasting 1 week or longer

    • Arthralgia lasting 2 weeks or longer

    • Typical rash (nonpruritic macular or maculopapular salmon colored over trunk or extremities while febrile)

    • Leukocytosis (10,000/mm 3 or greater with 80% or more neutrophils)

  • Minor Criteria

    • Sore throat

    • Lymphadenopathy and/or splenomegaly

    • Liver dysfunction

    • Negative tests for ANA and RF

  • Exclusions

    • Infections, malignancies, and rheumatic diseases

Diagnosis is made if there are five or more criteria, including at least two major criteria.

Adapted from Kumar et al.

Rheumatoid Factor–positive Polyarticular JIA. The presence of RF on two occasions at least 3 months apart was the criterion used to subcategorize polyarticular JIA into RF-positive and RF-negative disease. Anti-cyclic citrullinated peptide (anti-CCP) antibodies, a marker of this disease in adult rheumatoid arthritis (RA), had not been described at the time of the development of the criteria, and have not been considered in the revisions. As noted by Ferrell et al., anti-CCP–positive but RF-negative patients would therefore fail to meet ILAR criteria for RF-positive polyarticular JIA and would be classified as polyarticular JIA RF negative or oligoarticular JIA. The authors suggested that anti-CCP antibodies be included in any future revisions of the ILAR criteria, and proposed prioritizing serology (RF and anti-CCP) over the number of affected joints. According to that proposal, arthritis in any number of joints in the absence of systemic features (characteristic fever and rash), but with either RF or high titer anti-CCP on two occasions, would satisfy criteria for RF-positive JIA. Martini agrees that the number of affected joints should not be a criterion for classification. He suggests, further, that with the exception of antinuclear antibody (ANA)-positive arthritis, the other categories all represent childhood onsets of the same diseases that occur in adults, and that the term JIA should be abandoned.

The ILAR classification requirement for the presence of two positive tests for RF at least 3 months apart is not always fulfilled. A recent pilot study noted that children in an inception cohort who had only a single positive RF test rarely had polyarthritis. It is possible, therefore, that the requirement for two positive RF tests is an appropriate criterion for definition of RF-positive polyarthritis under the ILAR criteria.

ANA as a Classification Criterion. A number of studies have evaluated the presence of antinuclear antibodies as a classification criterion. Ravelli et al. concluded that patients in a group who were ANA positive (titer greater than or equal to 160 on rat liver and HEp-2 cell substrates on at least two occasions) shared many characteristics (early age at onset of arthritis, predominance of girls, asymmetric arthritis, risk of chronic anterior uveitis), irrespective of the number of inflamed joints, and that ANA positivity was a more appropriate classification criterion than the number of affected joints.

Juvenile Psoriatic Arthritis. Probably no category of the ILAR classification has been as difficult to deal with as JPsA. This stems from a number of factors: the onset of arthritis and psoriasis are not always synchronous, psoriasis is quite common in the general population, and the presence of psoriasis in a first-degree relative is difficult to verify. Manners has suggested that this criterion, together with the family history of a human leukocyte antigen (HLA) B27-associated disease in a first-degree relative be discarded for a number of reasons (unknown or uncertain family health status). Berntson et al. recommended eliminating the criterion of psoriasis in a second-degree relative. These recommendations require further testing.

Enthesitis-Related Arthritis. This category grew out of an understanding of the seronegative enthesitis and arthritis syndrome, which frequently eventuates in recognizable AS. Enthesitis has a central role in differentiating ERA from other categories of JIA. As critical as enthesitis is to the diagnosis of ERA, it may be difficult to diagnose clinically with certainty, and it must be noted that enthesitis occurs in other types of JIA, in particular psoriatic arthritis. The presence of imaging changes (sacroiliitis, lumbosacral spine disease) is not required for the diagnosis of ERA.

Undifferentiated JIA. The frequency of categorization of a child as having undifferentiated JIA has varied widely from around 13% to more than 50%, differences which almost certainly reflect differences in the application of the ILAR criteria.

After 20 years, the ILAR criteria are in need of revision; on that most pediatric rheumatologists are agreed. Just how and where the changes should be made remains controversial. Ideally, the application of genetic, transcriptomic, and proteomic analyses of children with chronic arthritis should provide the insight needed to make biologically (and therapeutically) meaningful changes.

Historical Review

The early descriptions of chronic arthritis in childhood are reviewed in Chapter 1 . Although not the first to recognize the disease, George Frederic Still presented the classic description of 22 children with chronic arthritis in 1897. He pointed out that the disease almost always began before the second dentition, was more frequent in girls, and was usually of insidious onset. Still observed that there was often no articular pain and that children exhibited a marked tendency to early contracture and muscle atrophy. The cervical spine was affected in the majority of cases, often during the early stages of the disease. The acute onset of disease in 12 patients who had lymphadenopathy, splenomegaly, and fever was described in detail. Pleuritis and pericarditis were common, although rash was not noted. Still suggested that childhood arthritis might have a different etiology from that of rheumatoid arthritis or that it might include more than one disease. Today, the term Still’s disease is most often used to describe the adult onset (beyond the age of 16 years) of this acute systemic arthritis. Although suggested by Still, recognition that chronic arthritis in children differed from that in adults took some time. It is now generally acknowledged that chronic arthritis in children differs clinically, often genetically and possibly pathogenically, from adult arthritis, with the exception of RF-positive polyarthritis, which is very similar to adult rheumatoid arthritis, and ERA, which may resemble AS as the disease progresses. Furthermore, the impact of JIA on the growing child is profoundly different from chronic arthritis in the adult.

The interested reader is referred to a number of case series of historic importance. These publications described experience with what today would be considered largely untreated patients, and they serve as a reminder of the serious potential long- and short-term effects of childhood arthritis, including death.


Incidence and Prevalence

JIA is not rare, but the true frequency is not known. It appears to be worldwide in distribution, but the reported incidence and prevalence vary considerably throughout the world. This may reflect the ethnicity, immunogenetic susceptibility, or environmental influences affecting the population under study, or may result from underreporting in the developing world, where data are very sparse and disease recognition and access to specialty care are limited (see Chapter 11 ). The reports of incidence and prevalence of chronic arthritis of childhood have been recently analyzed by Thierry et al. according to the classification criteria used. All but three reports described children from Europe or North America. Among the 33 prevalence studies, 11 used the EULAR criteria, 15 used the ACR criteria, and 7 used the ILAR criteria. The incidence rates ranged from 1.6 per 100,000 to 23 per 100,000, and the pooled incidence rate in Caucasians was 8.3 per 100,000 (confidence interval [CI] 8.1-8.7). Overall, pooled incidence rates were somewhat higher when the ILAR criteria were used (8.7 per 100,000), than with the ACR criteria (7.8 per 100,000) or EULAR criteria (8.3 per 100,000). Prevalence rates were evaluated in 29 papers (10 using ACR criteria, 11 using EULAR criteria, 7 using ILAR criteria, and 1 using ACR and EULAR criteria.) Prevalence rates ranged from 3.8 per 100,000 to 400 per 100,000, and the pooled prevalence rate was 32.6 per 100,000 (CI 31.3-33.9). Studies using the ACR criteria reported a higher pooled prevalence rate (45 per 100,000) than those using the EULAR criteria (12.8 per 100,000) or the ILAR criteria 30 per 100.000). Incidence rates were lowest in a Japanese study (less than 1 per 100,000), and highest in a Norwegian study (greater than 20 per/100,000). In the report of Danner and colleagues, the ILAR criteria were used in a hospital- or clinic-based population and demonstrated a prevalence of 19.8 cases per 100,000 children under 16 years of age. The difference between referred and community-based study populations in determining prevalence is emphasized by the survey by Manners and Diepeveen, in which a prevalence of 400 per 100,000 was found based on a physical examination by a pediatric rheumatologist of each of the 2241 12-year-old Australian schoolchildren included in the survey. Although this prevalence is considerably higher than that reported in most other studies, Mielants and colleagues found a prevalence of 167 per 100,000 in a similar study of Belgian children.

Estimates of the global frequency of JIA are shown in Table 15-2 . Using a range of reference data from European, North American, or Australian studies, an estimated 1.7 to 8.4 million children in the world have chronic arthritis, mostly undiagnosed.

TABLE 15-2

Estimates of the Global Incidence and Prevalence of JIA in Childhood

East Asia/Pacific 567,000,000 29,484-56,700 487,620-2,268,000
South Asia 536,000,000 27,872-53,600 444,880-2,144,000
Industrialized 225,000,000 11,700-22,500 186,750-900,000
Latin America 190,000,000 9880-19,000 157,000-760,000
Eastern Europe 138,000,000 7176-13,800 114,540-552,000
West/Central Africa 151,000,000 750-15,000 125,330-604,000
East/Southern Africa 147,000,000 7644-14,700 122,010-588,000
TOTAL 2,105,000,000 109,456-210,300 1,747,150-8,420,000

Incidence ranges and prevalence ranges are based on the range of published estimates.

Proportions of JIA Subtypes

In Europe and North America, oligoarthritis accounts for approximately 50% of patients with JIA in most large series. RF-positive polyarthritis is least common. In clinic-based studies from South Africa, India, and Zambia, oligoarthritis appears to be much less frequently recognized, and ERA and RF-negative polyarticular JIA are more commonly reported

Age at Onset

Quite distinct distributions of age at disease onset characterize each subset (see specific chapters). Disease onset before 6 months of age is distinctly unusual in any category of JIA, however.

Sex Ratio

In North and South America, Europe, and Australasia, twice as many girls as boys are affected. Marked differences in this ratio are apparent in the different onset types. In reports from South Africa, India, and Turkey, however, boys and girls are equally at risk, and in some instances, boys are more frequently affected. These differences may represent differences in the type of JIA that predominates in different areas of the world, or they may reflect biological differences or case ascertainment bias.

Geographical and Racial Distribution

The incidence and prevalence data outlined previously were derived primarily from American or northern European white populations. There are few comparable data for other geographical or racial groups. Nonetheless, suggestions of racial differences in frequency exist.

The incidence in Japan was reported to be low (0.83 per 100,000), and lower frequencies of chronic arthritis (identified as JRA) have been reported in children of Japanese, Filipino, or Samoan origin than in white children living in Hawaii. Chronic arthritis may be less common in North Americans of Chinese ancestry than in North American white children, but there is a paucity of data describing these children. Schwartz and colleagues concluded that the proportion of African-American children with JRA in a referral clinic population in the United States was consistent with their representation in the population served. However, there was a striking underrepresentation of this ethnic group in young children with oligoarticular and polyarticular onsets. Some reports suggest that chronic arthritis in children and adults is less frequent in African than in European populations. Saurenmann and colleagues compared the frequencies of JIA in children of differing ethnic origin in a multiethnic hospital-based cohort of more than 1000 children living in Toronto. They noted that children of African or Asian origin were underrepresented overall in comparison with their proportion of the healthy population. Children of European origin were more likely to have extended oligoarthritis or psoriatic arthritis than other ethnic groups. Those of Asian origin were more likely to have ERA, and those of African origin were more susceptible to RF-positive polyarticular disease. An analysis of white and First Nations children living in Western Canada suggests that, although the overall frequency of chronic arthritis in aboriginal children is not higher than that for the white population, the frequency of HLA-B27–associated arthritis is appreciably higher in the aboriginal group. Inuit children of northern Canada are reported to have a high overall incidence of chronic arthritis (23.6 per 100,000), and Boyer and associates suggested that the incidence of RF-positive polyarthritis is increased in southeast Alaskan Yupik Eskimo children. The numbers of patients in these studies are small, however, and conclusions with respect to actual incidence and prevalence of chronic arthritis in these groups are tentative.

Etiology and Pathogenesis

The etiology of JIA is unknown, although it is almost certainly multifactorial, and probably differs from one onset type to another. Systemic arthritis is characterized neither by the presence of autoantibodies nor a strong genetic predisposition and may be more appropriately considered to be an autoinflammatory disease. Autoantibodies are common in oligoarthritis (antinuclear antibodies) and RF-positive polyarthritis (IgM rheumatoid factor). In contrast, ERA, RF-negative polyarthritis, and, to some extent, psoriatic arthritis have less tendency to autoantibody formation but, in the case of ERA, are associated with misfolding of the protein coded for by the genetic marker HLA-B27. The strongest genetic associations with several categories of JIA remain those with genes coded for within the HLA or major histocompatibility complex (MHC) locus on chromosome 6, where a large cluster of immune related genes are found. Recent detailed, large genetic studies of JIA have confirmed these associations with HLA and also indicated several other susceptibility loci, which confirm the autoimmune etiology of JIA. That genetic factors are only part of the puzzle is demonstrated by the fact that, except for ERA and some cases of psoriatic arthritis, familial arthritis is very uncommon, although not unknown.

The fundamental pathological process is chronic inflammation, in which both the innate and adaptive immune systems play critical roles. In all categories of JIA, products of activated T cells and macrophages are involved in pathogenesis of synovitis.

Environmental influences have received little attention. One study suggested that breast-feeding has a protective effect on the development of JIA, especially in oligoarticular disease; however, a strong relationship was not confirmed in another investigation. Maternal smoking during pregnancy has been reported to be a risk factor for the development of arthritis in the first 7 years of life, especially in girls. The possible role played by epigenetic mechanisms is largely unexplored. The increasing understanding of the interactions between the immune system and the microbiome of the gut, and how they play a role in immune health or triggering of autoimmunity, has opened a new field that is likely to be of key importance to our understanding. In addition to polygenic genetic predispositions, disordered immune responses, and putative environmental triggers, any theory of pathogenesis must account for a number of factors: the clinical heterogeneity of the disease; the much higher prevalence of oligoarthritis, polyarthritis, and psoriatic arthritis in girls, and the strikingly higher incidence of ERA in boys; the narrow peak ages at onset for some types such as oligoarthritis, in contrast to the absence of a peak age at onset for systemic disease; and the association of extraarticular complications such as uveitis in certain disease subsets. There may be multiple etiological events, or the disorder may result from a single pathogenic vector with diverse clinical patterns evolving from interactions with the host. It may be postulated that an environmental agent affecting a child with a particular genetic predisposition, at a point of vulnerability—defined by age, intercurrent illness, prior antigenic experience, trauma, hormonal variations, psychological stress, or immunological maturity—results in a clinical disorder.

Immunopathogenic Mechanisms

A number of observations contribute to the hypothesis that the immune system is intimately involved in pathogenesis. First, there is abundant evidence of altered immunity, abnormal immunoregulation, cytokine production, and polymorphisms of genes involved in the immune response. Second, there is an association between specific immunodeficiencies and rheumatic diseases, including chronic arthritis (see Chapter 46 ). Third, there is a close relationship between immune reactivity and inflammation, the hallmark of arthritis. Whether it is principally an immunogenetically determined disorder or an antigen-driven immunological response, or the result of interactions between the two, is uncertain. The topics of immunopathogenesis and inflammation are discussed in detail in Chapters 3 and 4 , and in chapters discussing specific categories of JIA.

Hormonal Factors

The often striking differences in sex ratio, as well as the characteristic preadolescent or postadolescent peaks in incidence of specific categories of childhood arthritis, suggest that reproductive hormones may play important roles in pathogenesis. In one study, levels of progesterone and 17β-estriol were similar in patients with chronic arthritis and age-matched controls, but levels of dehydroepiandrosterone and testosterone were lower. Testosterone synovial fluid levels were lowest in those with disease of the longest duration. Low androgen levels may contribute to pathogenesis because androgens exert a protective effect against cartilage degradation.

Prolactin, produced by cells of the anterior pituitary and other cells, including lymphocytes, is elevated in inflammatory joint disease and inhibits chondrocyte apoptosis and related cartilage loss. Levels were increased in children with chronic arthritis and antinuclear antibody (ANA) positivity. The prolactin concentration correlated with levels of IL-6 and with a chronic disease course. Serum and urinary morning cortisol levels were somewhat diminished in children with active JIA.

Infection and Immunizations

Arthritis after viral infections is probably common, although it is usually self-limited. Arthritis in children has been linked to perinatal infection with the influenza virus A2H2N2, and to parvovirus B19. Chronic arthritis (especially RF-negative polyarthritis, spondyloarthritis, and oligoarthritis) has been described in children with human immunodeficiency virus infections. The cyclical pattern of incidence of chronic arthritis documented from 1979 to 1992 in Manitoba by Oen and colleagues correlated with the occurrence of infections to Mycoplasma pneumoniae.

The relationship between highly conserved bacterial heat shock proteins (HSPs) and chronic arthritis in animal models and humans has been extensively studied. Humoral and cellular immune responses to HSPs are present in children with chronic arthritis. HSPs have been demonstrated in the serum and synovial fluid. Van Eden and colleagues postulated that HSP-reactive T cells are part of the normal immune repertoire for TCR V-gene products; self-HSPs and bacterial HSPs may trigger this response. In 13 of 15 children with oligoarthritis, T-lymphocyte proliferative responses to HSP 60 were detected an average of 12 weeks before clinical remission of the inflammatory disease. Spontaneous remission was characterized by the presence of CD30 + T cells directed to HSP 60. Albani and associates demonstrated immune responses to the dnaJ HSP from Escherichia coli , especially in children with polyarticular disease. This protein has five amino acids that are homologous with those in the binding groove of DRB1, which in itself is increased in frequency in these children. Therefore, molecular mimicry may play a role in pathogenesis. Intranasal administration of a heat shock peptide suppresses adjuvant induced arthritis in mice. Despite circumstantial evidence, a direct link between infection and chronic childhood arthritis has remained elusive.

Postvaccination arthritis has been described after routine immu­nizations, and after measles-mumps-rubella vaccine, a persistent arthropathy was documented in one study, predominantly in females. However, a number of recent studies have demonstrated the safety and efficacy of vaccine administration to children with JIA ( Chapter 11 ).

Psychological Factors

High levels of psychological stress are common in families of children with arthritis. Psychological factors inherent in the family and child affect their adaptation to chronic illness, but their role in the causation of the disease is controversial. A recent study documented the occurrence of stressful life events preceding the onset of arthritis in many children. Some studies suggest that susceptibility to arthritis is associated with dysregulation of the autonomic nervous system that leads to an inappropriate response of the child’s immune system to stimuli and may be influenced by neuroendocrine gene polymorphisms.

Physical Trauma

Chronic arthritis has been reported by parents to follow minor physical trauma to an extremity. Such trauma may serve as a localizing factor, or it may simply call attention to an already inflamed and weakened joint. The fact that certain joints (e.g., the knee) are frequently affected in JIA could be interpreted to suggest that trauma associated with weight bearing in the young child is a factor in initiating chronic inflammation. However, there is no convincing evidence that trauma is a cause of JIA.

Nutritional Factors

To date, there has been no reproducible evidence that nutritional deficiencies or ingestion of particular foods either cause or exacerbate JIA. However, gluten ingestion in children with gluten sensitivity may exacerbate gut disease and as a consequence cause joint inflammation.

Animal Models of Inflammatory Joint Disease

The study of animal models of human disease provides clues about etiology, but these models are, at best, approximations of human disease. Some of these models are listed in Table 15-3 .

TABLE 15-3

Animal Models of Chronic Arthritis

Adjuvant Arthritis Mice Muramyl dipeptide T cell mediated
Collagen Arthritis Mice, rats Native type II collagen Antibody mediated
Infectious agents Many species Erysipelothrix, Mycoplasma, Chlamydia psittaci
Genetically determined MRL/1 mice
K/BxN mice
HLA B27 transfected mice, rats

Genetic Background

Familial JIA

Familial chronic arthritis is uncommon. Over a 10-year period, an American registry identified 200 sets of siblings identified as having JRA, of whom 21 are twins (13 identical). Within any one family, arthritis tends to have the same type of onset, and even the same complication of uveitis. In no instance was disease onset simultaneous, although in most cases the ages at onset were similar. There is no association between birth order and the development of arthritis.

The development of chronic arthritis in twins has been extensively studied. Concordance rates were 44% in identical twins and 4% in dizygotic twins. In a multicenter study of 71 affected sibling pairs and a trio of three siblings in three families, the mean interval between onset of disease in the siblings was 4.4 years. More than three quarters were concordant for onset type and disease course. Among seven sets of twins, the interval between disease onset was shorter (3.3 months) and all were concordant for onset type (six oligoarthritis, one polyarthritis). Uveitis was concordant in only 3 of 16 sibling pairs. There is also an increased prevalence of autoimmunity and of autoantibodies in first-degree relatives of children with chronic arthritis. Chronic inflammatory rheumatic diseases are more common in parents of multiple siblings affected by chronic arthritis.

One further association bears attention: the occurrence of JIA and adult RA in the same family. Documentation of this event is scant, and it must be concluded that with the exception of RF-positive polyarticular JIA, JIA and RA uncommonly occur in the same family.

Genetic Associations

Genetic influences on both JIA susceptibility and phenotype are polygenic. Genetic associations have been recently reviewed by Angeles-Han and Prahalad. Until recently, there were five well-established genetic risk factors: HLA class I and class II genes, the PTPN22 and PTPN2 genes, and the IL2RA/CD25 gene. Early studies revealed that specific HLA classes I and II alleles, and whole haplotypes, are associated with particular types of JIA. Some HLA associations parallel those observed in adult disease, such as HLA-DRB1*0401 in RF-positive polyarthritis, whereas others are particular to JIA subtypes such as the associations of HLADRB1*0801, 1101, and 13 with oligoarticular JIA. A recent international effort has led to the largest ever genetic study of JIA, with a focus on oligoarticular and RF-negative polyarticular JIA. This study increased the loci known to be associated with JIA by identifying 14 associated genes at genome-wide significance and a further 11 highly associated loci. Several other genetic associations are JIA subtype specific, such as with the endoplasmic reticulum aminopeptidase-1 (1) gene in ERA and IL-23R in psoriatic arthritis ; these associations are considered in subsequent chapters.

There may also be an association of the various types of arthritis with childhood onset with chromosomal abnormalities. Chronic JIA-like arthritis has been observed in patients with IgA deficiency, and with deletion of the short arm of chromosome 18 ( Chapter 46 ), 22q11.2 deletion syndrome (also called velocardiofacial or DiGeorge syndrome), and Turner syndrome. In a study of this association, 18 of approximately 500 children with JRA were found to have Turner syndrome (X0 chromosomes); polyarticular disease was present in 7.

It is not entirely clear whether the arthropathy of Down syndrome (trisomy of chromosome 21) differs from JIA. In a community-based survey, chronic arthritis was reported in 0.2% of 440 children with Down syndrome. In some children, the arthritis is more like psoriatic arthritis; and psoriasis is seemingly increased in frequency in children with trisomy 21. Of interest, both RF and anti-CCP antibodies are frequently found in children with Down syndrome even in the absence of arthritis. Arthritis was described in a child with partial trisomy 5q, monosomy 2p.

Clinical Manifestations of JIA

Constitutional Signs and Symptoms

Fatigue, anorexia, weight loss, and growth failure occur in many children with active JIA. Significant fatigue is rarely a feature of mild or limited articular disease, but it is a common symptom in children with polyarticular or systemic disease, especially at onset and during periods of poor disease control. Night pain may interrupt sleep and contribute to fatigue. Sleep fragmentation may also exacerbate pain as well as fatigue in these children. In a recent study, however, Ward and colleagues could not demonstrate a difference between children with arthritis and healthy subjects with respect to sleep characteristics. Anorexia may result from gastric irritation secondary to nonsteroidal antiinflammatory drug use, or nausea secondary to methotrexate use. Weight loss signifies the presence of active disease or an associated condition such as inflammatory bowel disease or celiac disease. Growth retardation is a frequent consequence of active JIA but is also be contributed to by prolonged corticosteroid use.

Puberty and secondary sexual characteristics are often delayed in children with active inflammation. In a group of Italian girls with JIA, menarche was later than in their mothers or in age-matched normal controls.

Musculoskeletal Manifestations

Pain and Stiffness

Aspects of pain in children with arthritis have been reviewed by Anthony and Schanberg and by LaLouviére et al. It is the rare child who has no pain in the presence of active arthritis. Although a child with chronic arthritis may not complain of pain at rest, active or passive motion of a joint elicits pain in the inflamed joint, particularly at the extremes of the range of motion. Pain is usually described as aching or stretching and is of mild-to-moderate severity. In contrast, children with pain-amplification syndromes almost always describe pain as extremely severe (see Chapters 8 and 52 ). Pain elicited by pressure (tenderness) is usually maximal at the joint line; over hypertrophied, inflamed synovium; or at entheses. Bone pain or tenderness is not characteristic of arthritis, and its presence should alert the examiner to the possibility of a malignancy or infection involving bone, or chronic nonbacterial osteomyelitis.

Joint stiffness may occur, particularly on arising or after prolonged inactivity. More often, however, evidence of stiffness is provided by the parent, who describes slowness or awkwardness of the gait, most marked in the morning or after a nap or prolonged sitting, which improves with activity or the application of heat to the affected area. The significance of the duration of stiffness has not been extensively examined, but stiffness lasting more than 15 minutes signifies a considerable level of joint inflammation. Stiffness that lasts all day is suggestive of a pain amplification syndrome. The pathophysiology of joint stiffness is unclear.

Characteristics of the Inflamed Joint

An actively inflamed joint exhibits the cardinal signs of inflammation: swelling, pain, heat, loss of function, and sometimes erythema ( Fig. 15-1 ). Swelling of a joint may result from periarticular soft tissue edema, from intraarticular effusion, or from hypertrophy of the synovial membrane. Large synovial cysts are an unusual complication. They may occur in the antecubital area or anterior to the shoulder ( Fig. 15-2 ).When they occur in the popliteal space (Baker cyst) ( Fig. 15-3 ), they may rupture into adjacent muscles, and dissect into the calf. This event is characterized by sudden sharp pain and swelling in the calf, followed by crescentic ecchymoses about the malleoli. They may be the initial or sole manifestation of chronic arthritis and, if unilateral, may be misinterpreted as a tumor or as deep venous thrombosis. A healthy child may occasionally develop a transient popliteal cyst. Ultrasound imaging or magnetic resonance imaging (MRI) aids in making the correct diagnosis. Inflamed joints in children with JIA are often warm but almost never erythematous. In contrast, the joint may be erythematous in septic arthritis or acute rheumatic fever and in some of the other reactive arthritides.


The joints of the wrists and hands of a -year-old boy with systemic-onset disease are swollen, warm, and painful. The proximal and distal interphalangeal joints are erythematous. There are flexion contractures of the fingers.


Brachial synovial cyst. A 6-year-old girl developed this dissecting cyst of her right arm as the first manifestation of chronic arthritis. Later, bilateral effusions developed in both knees.


A, Popliteal cyst or Baker cyst. The cyst was associated with pain at the back of the knee, was somewhat tender, and transilluminated. Aspiration yielded clear yellow fluid of low inflammatory activity. B, Arthrogram of a popliteal cyst, with contrast medium outlining the communication ( arrow ) between the synovial space and this dissecting cyst in an 18-year-old boy who had had arthritis since the age of 9 years.

Inflamed joints usually lose range of motion, particularly in extension, because this is usually a position of relative comfort, accommodating shortened soft tissue structures, although to some extent this depends on which joint is involved. Hyperextension is characteristically the first range to be lost in the cervical spine. Elbows, wrists, metacarpophalangeal and interphalangeal joints usually lose range in extension or hyperextension first, but flexion range is also often diminished. The hallmark of intraarticular hip joint disease is loss of internal rotation and flexion. There is loss of hyperextension or extension, and sometimes flexion in the knees. Loss of ankle dorsiflexion, subtalar eversion, and midfoot supination are characteristic of an inflamed joint. Loss of dorsiflexion of the first metatarsophalangeal joint may be of considerable functional significance.

The Pediatric Gait, Arms, Legs, Spine screen is an excellent tool for the quick evaluation of the child for evidence of musculoskeletal disease. It is most applicable for use by the general physician or pediatrician and does not include the detailed examination that a pediatric rheumatologist would perform.

Distribution of Affected Joints

Any joint may be affected, but large joints (knees, ankles, wrists, elbows, hips) are most frequently involved. Small joints of the hands and feet may also be affected, particularly in polyarticular-onset disease. Disease in the apophyseal joints of the cervical spine occurs at onset in approximately 2% of children and may present as a torticollis. In the pre-methotrexate, prebiologics era, approximately 60% of patients eventually developed involvement of the cervical spine. Instability of this area may occur early, rendering the child at risk for injury in an accident or with attempted intubation before general anesthesia. The sternoclavicular, acromioclavicular, and sternomanubrial joints are infrequently affected. Cricoarytenoid arthritis is unusual but may be responsible for acute airway obstruction. Inflammation of the synovial joints of the middle ear—the incudomalleal and incudostapedial articulations—is rarely appreciated clinically. Tympanometric studies, however, have indicated that subclinical disease may be present in almost two thirds of children with JRA.

Patterns of joint involvement are often quite characteristic. Thus, symmetrical involvement of large and small joints is typical of polyarticular disease. Arthritis predominantly affecting the joints of the lower extremity characterizes ERA. The presence of hip joint disease is not uncommon in ERA, but rarely occurs in oligoarticular JIA. Psoriatic arthritis tends to be somewhat asymmetrical and involves both large and small joints, sometimes including the distal interphalangeal joints.


Tenosynovitis is quite common, but it is generally not a striking or isolated clinical complaint. The most common sites are the extensor tendon sheaths on the dorsum of the hand, the extensor sheaths over the dorsum of the foot, and those of the posterior tibial tendon and the peroneus longus and brevis tendons. If tenosynovitis is very prominent, Blau syndrome should be considered (see Chapter 39 ). Triggering or loss of extension of the fingers may result from a stenosing synovitis of the flexor tendon sheaths. Clinically recognized carpal tunnel syndrome is uncommon in children. Tenosynovitis of the superior oblique tendon of the eye may cause pain on upward gaze, sometimes with diplopia (Brown syndrome).


Bursae are synovial-lined sacs that do not usually communicate with the joint but, when inflamed, may be confused with intraarticular pathology because of their juxtaarticular locations. Trochanteric bursitis, olecranon bursitis, anserine bursitis (between the gracilis and semitendinosus tendons and the tibial collateral ligament), prepatellar bursitis, infrapatellar bursitis, and retrocalcaneal bursitis are typical sites of painful swelling. The Baker cyst, in the popliteal fossa, is usually an expansion of the preexisting gastrocnemius semimembranosus bursa, but in the child with an inflamed knee joint, it may communicate with the intraarticular space or represent a posterior extension of the synovium of the knee. Occasionally, bursitis develops in children from repeated trauma but no underlying rheumatic disease.


Osteopenia is a potentially major determinant of functional outcome in young adults who have had chronic arthritis as children (see Chapter 53 ). Children with chronic arthritis have a diminished bone mass and are at increased risk for fractures in adulthood and for an earlier onset of osteoporosis. Contributing factors include physical inactivity, high levels of inflammatory cytokines, poor nutrition, and low levels of vitamin D. An important determinant of future fracture risk is the peak bone mass achieved at the end of skeletal maturation, which is almost complete by the late years of adolescence.

Muscle Disease

Atrophy and weakness of muscles around inflamed joints occur frequently and are often accompanied by a shortening of the muscles and tendons that results in flexion contractures. Atrophy of the vastus medialis muscle is characteristic of arthritis in the knee. The calf muscles become atrophic when the ankle joint is involved. Sacroiliitis may be associated with atrophy of gluteal and thigh muscles. Muscles of the forearm become atrophic in children with chronic wrist arthritis. A few children with widespread, prolonged, active arthritis develop progressive muscle atrophy that is most severe and persistent if it occurs before 3 years of age.

Extraarticular Manifestations

Generalized Abnormalities of Growth

Abnormalities of physical growth and development may complicate chronic arthritis. Linear growth is retarded during periods of active systemic disease. Mean weight for age and weight per height are significantly diminished in those with polyarticular disease. Growth retardation (height less than 5th percentile) was documented in 50% of adults who had childhood onset of systemic JIA, 11% of those with oligoarticular onset, and 16% of those with polyarticular onset. Glucocorticoid therapy could only partly explain these observations. In a more recent study, a marked decrease in height velocity (−2 SD) was observed in 56% of children with systemic arthritis. In this study, growth retardation was associated with glucocorticoid administration, nutritional status, bone mineral density, and early disease onset. Each of these studies reflects the effect of disease in a population treated in the pre-methotrexate, pre-biologics era and is an important reminder of the potentially devastating effects of untreated or undertreated JIA on normal growth and development. With better disease control, severe growth retardation is increasingly uncommon. In a recent study of children with oligoarthritis, however, Padeh et al. noted generalized growth retardation in one third.

The mechanisms of growth suppression in JIA involve high levels of proinflammatory cytokines, such as IL-6, IL-1b, and tumor necrosis factor (TNF)-α, seen in children with active arthritis that directly or indirectly influence growth plate chondrocytes and linear bone growth. Levels of growth hormone and IGF-I and -II may be reduced. IGF-I was inversely correlated with IL-6 levels in children with systemic disease.

Localized Growth Disturbances

During early active disease, development of the ossification centers is accelerated, apparently related to the hyperemia of inflammation and local production of growth factors. The result may be either overgrowth of the affected limb or, ultimately (though much less commonly), premature fusion of the involved physes, resulting in diminished length.

If arthritis occurs in one knee only, a discrepancy of leg lengths results. A difference of greater than 1 to 2 cm is probably functionally significant, and differences of 5 cm or more occasionally occur. Apparent leg-length inequality may also result from pelvic rotation and scoliosis. As the child grows, inequalities of minimal to moderate degree may disappear, but they persist in up to two thirds of these children. Significant leg-length inequality was reported to occur much less frequently in a setting in which intraarticular corticosteroids were used to treat inflammation.

Micrognathia and/or retrognathia may result from growth disturbances of the mandible as a consequence of arthritis in the temporomandibular joint (TMJ). Early TMJ arthritis is difficult to detect clinically and often not until growth changes are evident. Many children with MRI-confirmed TMJ arthritis do not complain of pain, although some have pain on opening the mouth or experience clicking on opening or closing the mouth. In a Swedish survey of 70 children with JIA, 56% had symptoms (crepitus, pain, difficulty opening the mouth) and 41% had radiographic evidence of TMJ pathology attributable to arthritis. In one patient, the disease began in a TMJ. Cannizzarro et al. found that 38.6% of 223 children with JIA had involvement of the TMJ a mean of 4.6 years after disease onset, based on clinical and radiographic evidence. Children with extended oligoarticular JIA were most commonly affected (61%), followed by those with RF-negative JIA (52%), psoriatic JIA (50%), systemic JIA (36%), RF-negative polyarticular JIA, persistent oligoarticular JIA (33%), and ERA (11%). A younger age at onset, a high erythrocyte sedimentation rate (ESR) at diagnosis, upper extremity involvement, and the absence of HLA-B27 were significantly associated with the development of TMJ arthritis. Extreme micrognathia is most likely to occur if arthritis begins before 4 years of age and is poorly controlled.

The mandible ossifies by intramembranous bone production. A number of factors contribute to mandibular growth abnormalities. Pain in the TMJ may inhibit normal masseter muscle development, which in turn retards mandibular bone development, resulting in a shortened mandibular ramus and body. Destruction of the condyle of the mandible causes further diminution of overall mandibular height. In some children, overgrowth of the condyle may contribute to TMJ dysfunction. As demonstrated by MRI, the disease is bilateral in three quarters, but early disease may appear to be unilateral on clinical examination. It is characterized by mandibular asymmetry; deviation to the affected side on opening of the jaw; difficulty in palpating the affected mandibular condyle; and pain, tenderness, or crepitus of one or both TMJs. Heterotopic bone formation has been described in 12 children with severe TMJ arthritis. A report of routine MRI evaluations of the TMJ in children with JIA suggests a high frequency of asymptomatic disease and describes protocols for assessing this joint. Little is understood about the effect of chronic arthritis on dental caries or periodontal disease.


There persists a widespread but unsubstantiated assumption by parents and some health professionals that physical activity in children with arthritis should be limited. Children with arthritis tend to be less physically active than their peers. Lelieveld and colleagues documented limited physical activity in adolescents with JIA compared to their peers, as measured by a 3-day activity diary and concluded that adolescents with arthritis have low aerobic and anaerobic fitness. Takken and colleagues noted reduced maximal oxygen consumption in children with JIA, compared with controls.

Skin and Subcutaneous Tissue

The classic rash of systemic-onset disease is discussed in Chapter 16 . A second cutaneous change, occurring particularly in children with involvement of the hands, is a dark discoloration of the skin over the proximal interphalangeal (PIP) joints. The presence of this finding may reflect disease chronicity. In children with tender joints, retention keratosis may simulate a pigmented lesion and reflects the inability to perform adequate self-care. Nodules occur in a variety of rheumatic diseases in children. Subcutaneous rheumatoid nodules occur in 5% to 10% of children with chronic arthritis, almost always confined to those with RF-positive disease (see Chapter 17 ). They most commonly occur over the proximal ulna and the Achilles tendon. Tendon-associated nodules may reflect tenosynovitis, especially in the flexor tendons of the fingers.

Asymmetrical lymphedema of the subcutaneous tissues of one or more extremities has been documented in several children with arthritis. The swelling is usually painless and may be pitting. The cause is unknown but does not seem to be related to local obstruction caused by joint swelling. The course is chronic but may improve over several years. Cutaneous vasculitis is very rare and occurs most often in the older child with RF-positive polyarthritis.

Ocular Disease

Ocular inflammation may occur at any time in the course of JIA (see Chapter 22 ). Uveitis is characteristically asymptomatic, except in ERA, where it is usually characterized by a painful pink eye. Reactive arthritis is also associated with symptomatic conjunctivitis and occasionally uveitis. Keratoconjunctivitis sicca occasionally occurs, particularly in RF-positive polyarthritis (see Chapter 30 ). Scleritis is rare.


The histopathological features of synovitis in patients with JIA are similar to those described in RA. There is villous hypertrophy and hyperplasia of the synovial lining layer ( Fig. 15-4 ). The subsynovial tissues are hyperemic, edematous, and show a dense infiltration by T and B lymphocytes, plasma cells, macrophages, and dendritic cells, and in some cases natural killer (NK) cells. Endothelial hyperplasia is often prominent, likely driven by production of the angiogenic chemokines, and pro-angiogenic factors including vascular endothelial growth factor (VEGF) and osteopontin, whose production is increased in hypoxic conditions. There is a selective accumulation in the synovium of activated T cells, which are clustered around antigen-presenting dendritic cells. Myeloid dendritic cells (mDC) are typically localized to the synovial lining layer, whereas interferon-α secreting plasmacytoid dendritic cells (pDC) are in T and B cell rich aggregates. Both CD4 and CD8 synovial T cells are highly activated. These T cells produce abundant cytokines including tumor necrosis factor (TNF), interferon (IFN)-γ and, as more recently recognized, IL-17 and granulocyte macrophage colony-stimulating factor (GMCSF). In parallel, synovial monocytes and tissue macrophages produce cytokines and chemokines that both contribute to local cartilage and bone damage and also to the recruitment of more inflammatory cells. An exuberant synovitis eventually results in progressive erosion and destruction of articular cartilage and, later, of contiguous bone with pannus formation.

Jun 30, 2019 | Posted by in RHEUMATOLOGY | Comments Off on Juvenile Idiopathic Arthritis
Premium Wordpress Themes by UFO Themes