Chronic childhood arthritis affecting five joints or more during the first 6 months of disease is defined as polyarthritis. The International League of Associations for Rheumatology (ILAR) classification system for juvenile idiopathic arthritis (JIA) further categorizes polyarthritis as rheumatoid factor (RF) negative if tests for RF are negative, and RF positive if RF is detected on two occasions at least 3 months apart ( Table 17-1 ). RF-negative and RF-positive JIA subsets have distinguishing clinical features, disease courses, and outcomes; therefore, in this chapter RF-negative and RF-positive polyarthritis JIA subsets are considered separately as distinct clinical entities.
|Arthritis affecting five or more joints during the first 6 months of disease; a test for RF is negative |
|Arthritis affecting five or more joints during the first 6 months of disease; two or more tests for RF at least 3 months apart during the first 6 months of disease are positive |
Rheumatoid Factor Negative Polyarticular JIA
Polyarthritis accounts for approximately 20% of JIA patients; of these, approximately 85% have negative tests for RF, although RF frequencies vary in accord with ethnicities.
Incidence and Prevalence
Incidence and prevalence data vary widely because of differences in case ascertainment, diagnostic and classification criteria applied, accessibility to care, and genetic and ethnic characteristics of the respective populations. The estimated incidence of chronic childhood arthritis varies from 7 to 21 per 100,000 in North American and Northern European studies. Prevalence rates of 121 to 220 per 100,000 have been reported; meta-analysis indicates a chronic childhood arthritis prevalence of 132 (95% confidence interval [CI] 119 to 145) per 100,000 from population studies. Estimating that 20% of JIA populations have polyarthritis and 85% of the polyarthritis populations are RF-negative, the annual incidence and prevalence for RF-negative polyarthritis can be estimated as 1 to 4 per 100,000 and 21 to 37 per 100,000, respectively.
Age at Onset and Sex Ratio
RF-negative polyarticular JIA can begin at any age before 16 years, but onset age displays a biphasic trend with a peak at ages 1 to 3 years and another encompassing later childhood and adolescence. RF-negative polyarthritis affects girls four times more frequently than boys. The predominance of females is greater in those with an onset during adolescence (female-to-male ratio, 10 : 1) compared with those with a younger onset age (female-to-male ratio, 3 : 1). Younger onset RF-negative polyarthritis patients are more likely to be antinuclear antibody (ANA) positive and are at greater risk for iridocyclitis. A young onset age of RF-negative polyarthritis is associated with a less favorable long-term outcome. Greenwald et al. found 2-year outcomes in an older onset, RF-negative JIA subgroup to be generally favorable even prior to the era of biological therapies.
Geographic and Racial Distribution
JIA occurs worldwide but prevalence varies widely among geographic regions. Oen and Cheang reported that polyarthritis accounted for a higher proportion of East Indian (61%) and North American Indian (64%) children with chronic arthritis, compared with white children (27%). Saurenmann et al. analyzed ethnicity as a risk factor for JIA in a multiethnic cohort; among 223 children with RF-negative polyarthritis, no significant differences among European and non-European patients were found. However, the North American Indian population had a high relative risk (3.2) of developing RF-negative polyarthritis.
Etiology and Pathogenesis
The etiologies of the respective JIA subtypes are unknown but are thought to have complex origins that include interactions among an array of susceptibility genes and as yet unidentified exogenous factors. Environmental and lifestyle influences have been proposed as factors promoting arthritis in the context of genetic vulnerability.
There are no cytokine or chemokine response patterns in either blood or synovial fluid that are unique to RF-negative polyarthritis. De Jager and colleagues noted comparable plasma level increases in interleukin (IL)-6 and -12, and chemokines C-C motif ligand (CCL3), C-X-C motif ligand (CXCL)9, and CXCL10 were found in a small group of children with RF-negative polyarthritis (10 patients) and oligoarthritis with a polyarticular course (5 patients). Increased levels of IL-17 found in seronegative polyarthritis are considered to be of potential pathogenic importance because IL-17 promotes other proinflammatory cytokines and enhances matrix metalloproteinase production, leading to cartilage degradation. CCL20 derived from synovial fluid mononuclear cells was increased in children with polyarthritis (including those with extended oligoarticular JIA); the enhanced production was attributed to the hypoxic synovial environment. The hypoxic synovial environment has also been suggested as a factor that promotes increases in intraarticular vascular endothelial growth factor and osteopontin, which enhance angiogenesis in synovial tissue in children with RF-negative polyarthritis and extended oligoarthritis JIA subsets.
A genetic influence in the pathogenesis of JIA is indicated by ethnic variability in the incidence of certain JIA subsets, female preponderance, increased sibling recurrence rates, and associations with both human leukocyte antigen (HLA) and non-HLA genes.
Genes both within and outside the major histocompatibility complex (MHC) contribute to genetic susceptibility to JIA. The HLA class I A2 allele confers susceptibility in RF-negative polyarthritis as do the class II alleles DRB1*08, DQAI*04, and DPB1*03. These HLA-related profiles are distinct from those characterizing RF-positive polyarthritis, supporting the view that children without RF have a disease different, at least genetically, from children with RF. Further, certain HLA alleles that confer susceptibility to RF-negative polyarthritis (A2, DRB1*08, and DQA1*04) also confer susceptibility to the oligoarthritis JIA subtype, suggesting that RF-negative polyarthritis is more allied genetically with oligoarticular JIA than with RF-positive polyarthritis. The association of HLA-DRB1 is similar between oligoarticular JIA and the younger onset subgroup of polyarticular JIA; the HLA-DRB1*1103/1104 haplotype is found in the oligoarticular cohort and the younger polyarticular group in contrast with the HLA-DRB1*08 susceptibility haplotype found in the older polyarticular JIA group.
In the MHC of children with oligoarticular JIA and, to a somewhat lesser extent with RF-negative polyarticular JIA, a single nucleotide polymorphism (SNP), rs7775055, which tags the HLA-DRB1*0801–HLA-DQA1*0401–HLA-DQB1*0402 haplotypes, is frequently found, again suggesting these two JIA subsets tend to share common genetic influences.
The TRAF1/C5 region on chromosome 9 encodes the tumor necrosis factor (TNF)-receptor-associated factor 1 and the complement component 5. In polyarticular JIA there is an increase in the A allele of an SNP in the TRAF1/C5 region when compared with controls.
The protein tyrosine phosphatase nonreceptor type 22 (PTPN22) gene codes for lymphoid-specific phosphatase, which modulates antibody-mediated T-cell activation. A missense SNP in the PTPN22 gene reduces the ability to downregulate T-cell activation and has been associated with JIA. There is an association of this PTPN22 SNP and RF-negative polyarthritis ; however, the association has not been found consistently, possibly reflecting ethnic differences in the populations studied. Certain PTPN22 SNPs, although found with significant frequency in RF-negative polyarthritis, are not specific for this JIA subtype, as the same markers are found in RF-positive and oligoarticular JIA.
Continued mining of JIA genome-wide scans is expected to expose additional susceptibility genes allowing for more precise biologically based categorization of JIA, providing insight into mechanisms of disease and informing new treatment strategies.
In children with RF-negative polyarticular JIA, joint disease predominates; extraarticular features are infrequent. Variations in onset ages, clinical and serological features, and courses suggest the RF-negative class of polyarthritis comprises different clinical entities. For example, some patients with RF-negative polyarthritis have a young onset age, positive tests for ANA and uveitis, and therefore, apart from the number of involved joints, are similar to the oligoarthritis JIA subtype.
Onset of arthritis may be acute, but it is more often insidious, with progressive accumulation of additional joints. Morning stiffness, indicative of active arthritis, may persist for hours or occasionally all day. The arthritis may be remittent or indolent. Joint swelling is a result of synovial hypertrophy and/or intraarticular fluid. Joints may be warm but are generally not tender or red. Hot, red, tender joints suggest infection or malignancy rather than JIA. Among children with RF-negative polyarthritis knees, wrists and ankles are most commonly affected. Small joint involvement of the hands or feet may occur early or late in the course of the disease; distal interphalangeal joints are seldom affected at onset ( Fig. 17-1 ). The temporomandibular joint (TMJ) is commonly affected in children with a polyarticular disease course regardless of onset subtype, but those with RF-negative polyarthritis are more likely to have TMJ involvement, particularly at long-term follow-up, than those who are RF-positive. The earlier age of onset, when the TMJ might be more vulnerable to damage, is thought to be a reason for the greater prevalence of radiographically evident destruction in the RF-negative group. Advances in TMJ imaging, including computed tomography and magnetic resonance imaging (MRI), demonstrate that TMJ arthritis is more common and the severity more variable than can be discerned by clinical assessment alone; TMJ arthritis can be present even in the absence of clinical symptoms or signs. In experienced hands ultrasonography can be a suitable method for detecting and monitoring TMJ arthritis. Joint damage associated with TMJ arthritis can result in altered physiognomy, including micrognathia, retrognathia, malocclusion, and facial asymmetry ( Fig. 17-2 ).
Cervical spine involvement is not commonly recognized early in the course of RF-negative polyarthritis either clinically or by conventional radiography but with longer term follow-up decreased range of motion can ensue. More sensitive imaging modalities, such as MRI, reveal that cervical spine involvement is more common than is generally appreciated even in asymptomatic patients. Atlantoaxial subluxation ( Fig. 17-3 ), erosion of the odontoid process, and vertebral ankylosis can occur ( Fig. 17-4 ).
In RF-negative polyarthritis the number of affected joints tends to be less and the pattern of involvement more asymmetric than in RF-positive polyarthritis ( Fig. 17-1 ). In RF-negative disease involvement of wrists and small joints of the hands is less frequent in RF-positive disease. Clinical signs of hip involvement are present in fewer than 20% with RF-negative polyarthritis at first presentation, but progressive hip abnormalities become evident with longer term follow-up. Oen et al. reported that radiological signs of hip joint involvement during childhood were more likely to occur in RF-negative than RF-positive polyarthritis; the tendency for RF-negative polyarthritis to have its onset at a younger age compared with RF-positive disease might be a factor contributing to the higher frequency of abnormalities identified during childhood in the seronegative group. Wrist and ankle involvement at first presentation has been suggested as a predictor of progression to polyarthritis in those first presenting with oligoarthritis.
A small subset of RF-negative children have “dry synovitis,” a condition characterized by polyarthropathy, minimal or absent clinical signs of joint effusion or synovial hypertrophy, joint stiffness, limited range of motion, and joint contractures associated with laboratory indicators of inflammation. This uncommon subset of polyarthritis can be considered a variant of RF-negative polyarthritis although it has also been suggested to be a forme fruste of scleroderma.
Systemic manifestations in seronegative polyarticular JIA are unusual but can include fatigue and growth failure. Fever seldom occurs and when present is low grade.
Fatigue in children with polyarthritis can be present even in the absence of active joint disease. Ringold et al. studied fatigue in 60 children with polyarthritis, of whom 24 (61.2%) were RF negative, using a survey tool that includes assessment of general fatigue, sleep/rest fatigue, and cognitive fatigue domains. Both JIA children and their parent/proxies reported lower scores in all domains compared with controls. Fatigue in JIA is significantly associated with poor functional ability. Factors contributing to fatigue can include pain and stress, decreased muscle mass, low aerobic and anaerobic capacity, and anemia. Sleep disturbance does not appear to be a common factor contributing to fatigue in JIA.
Growth disturbances are common in JIA. In children with polyarthritis height-for-age Z scores (reflecting the number of standard deviations from the normal mean) may decline in the first several years of disease but tend to return to normal. In those children without RF, the negative deviation is less marked and less prolonged than those with RF, particularly with longer disease duration. Low growth velocity tends to correlate with disease severity and activity and with the number of involved joints.
Subcutaneous nodules are rare in RF-negative polyarthritis. Among 131 children with RF-negative polyarticular JIA from a Canadian JIA inception cohort, only 1 (0.7%) had subcutaneous nodules at initial presentation.
Next to oligoarticular JIA (which accounts for more than half of JIA patients affected by uveitis), chronic asymptomatic uveitis is most common in the RF-negative polyarthritis group. Approximately 15% to 20% of children with RF-negative polyarthritis have uveitis and account for approximately 20% of all JIA uveitis patients.
Sabri and colleagues reported that 32 of 142 JIA patients (22.5%) with uveitis had RF-negative polyarthritis; none with RF-positive JIA had uveitis. As with the oligoarthritis, JIA subgroup uveitis in RF-negative polyarthritis tends to be associated with younger onset age and ANA positivity. Greenwald et al. reported that only 2% of patients with older onset RF-negative polyarticular JIA had uveitis despite the fact that 47% were ANA positive.
Cardiovascular and pulmonary disease.
RF-negative polyarthritis is not typically associated with overt cardiovascular pathology. Bharti et al. reported that children with arthritis, regardless of onset subtype, had significantly greater left ventricular volumes and other abnormalities suggesting abnormal left ventricular diastolic relaxation. Knook et al. demonstrated lower 1-second forced vital capacity and peak expiratory flows in a group of 31 children with chronic arthritis, of whom more than two thirds had RF-negative polyarthritis. These abnormalities were attributable to impaired respiratory muscle strength rather than intrinsic restrictive or obstructive lung disease.
The differential diagnosis for a child with polyarthritis includes other rheumatic diseases, infections, other inflammatory conditions, malignancies, and metabolic and genetic disorders.
The onset of polyarthritis in a girl later in childhood or during adolescence could suggest the possible diagnosis of systemic lupus erythematosus (SLE). The arthritis of SLE may mimic that of JIA, although it is nonerosive and less likely to be deforming; the presence of other clinical hallmarks and a positive test for anti-double stranded DNA (anti-dsDNA) antibodies establishes the diagnosis of SLE. Ragsdale and colleagues described nine children (six of whom were RF negative) who developed SLE years after an initial diagnosis of polyarticular juvenile arthritis.
The differential diagnosis of RF-negative polyarthritis also includes enthesitis-related arthritis (ERA). Predominant involvement of large joints of the lower extremities and the presence of enthesitis supports the diagnosis of ERA although enthesitis can occur, albeit uncommonly, in other types of JIA.
Scleroderma begins insidiously with joint contractures of the small joints of the hands mimicking features of polyarthritis but ordinarily without associated signs of intraarticular swelling. Children with dermatomyositis may have polyarthritis, but they can be distinguished from those with JIA by clinical manifestations such as typical rash and muscle weakness.
Septic arthritis affecting multiple joints is unusual; only 3% of the 65 children with septic arthritis reported by Al Saadi et al. had more than one involved joint. Lyme disease may be polyarticular, but it can usually be differentiated from RF-negative polyarthritis by its intermittent pattern of arthritis activity and accompanying extraarticular abnormalities. Arthritis caused by Neisseria gonorrhoeae may have an early migratory polyarticular phase.
Reactive polyarthritis in response to infection in the respiratory, gastrointestinal, or genitourinary tracts ordinarily can be distinguished from polyarticular JIA by a limited duration of the disease and associated clinical manifestations. Acute rheumatic fever following group A beta-hemolytic streptococcal pharyngitis is characterized by acute, painful, nonerosive, migratory polyarthritis (see Chapters 43 and 44 ).
Malignant infiltration of bone or synovium can mimic polyarthritis, although in most instances the malignant focus is in juxtaarticular bone rather than in the joint. However, joint swelling can occur in lymphoblastic leukemia as a result of leukemic infiltration of the synovium. Joint involvement in malignancy tends to be oligoarticular rather than polyarticular and associated with systemic manifestations of malignancy.
Other Inflammatory Conditions
Arthritis associated with inflammatory bowel disease or sarcoidosis should be considered in the differential diagnosis of RF-negative polyarthritis. Sickle cell disease causes diffuse, symmetrical swelling of the hands and feet that may mimic true arthritis. Hypermobility syndromes, mucopolysaccharidosies, familial hypertrophic synovitis, familial arthritis and camptodactyly, familial osteochondritis dissecans, Stickler’s syndrome, velocardial facial syndrome, Turner’s syndrome, and relapsing polychondritis are rare causes of disease that may suggest polyarticular JIA. Scurvy should be considered in the differential diagnosis of polyarthropathy in children at risk for nutritional deficiencies, including those with autism and other developmental disorders.
Laboratory tests can provide evidence of inflammation, are useful in excluding other diagnoses, and important in classification, prognostication, and guiding therapy (see Chapter 10 ).
Indicators of Inflammation
Children with polyarthritis typically have moderate elevations of the erythrocyte sedimentation rate (ESR) and C-reactive protein, and may have elevated white blood cell and platelet counts and a normocytic, hypochromic anemia associated with chronic inflammation.
Applying ILAR criteria, discrimination between RF-positive and RF-negative polyarthritis is based on at least two positive RF results as determined by an accredited laboratory. RFs are antibodies that bind to the CH2 and CH3 domains of the Fc portion of IgG. Customarily, RF is detected by agglutination assays that preferentially detect pentameric IgM RF. Approximately one third of children with polyarthritis who do not have IgM RF detectable by agglutination methods have IgM RF detected by more sensitive enzyme immunoassays (EIAs). RF detected by either technique is associated with deforming and erosive joint disease. Further, children with IgM RF-negative polyarthritis, determined by conventional methods, can have “hidden RFs.” Hidden IgM RF is 19S IgM RF that, because it is bound to IgG in the serum being tested, cannot generate a response in a standard agglutination assay until it is acid eluted from the IgG. Up to 85% of children with polyarticular disease have been reported to have such antibodies associated with active disease. IgA RF, alone or in combination with IgM RF, has been associated with active disease, disability, and radiographic joint space narrowing and bone erosions in polyarthritis.
Anticitrullinated peptide/protein antibodies (ACPA) have been reported to occur in 0% to 17% of children with RF-negative polyarthritis.
ANAs are present in approximately half of children with RF-negative polyarthritis. The group of ANA-positive RF-negative polyarthritis patients is generally not substantially different from the group with oligoarthritis with respect to age at first presentation, sex ratio, or prevalence of uveitis, suggesting that ANA positivity, irrespective of JIA onset subtype, distinguishes a relatively homogeneous group characterized by early onset, female predominance, asymmetric arthritis, and risk of uveitis. The antigenic specificities of ANA in JIA are generally unknown ; antibodies to individual histones and to histone-histone and histone-DNA complexes are occasionally, but inconsistently, found.
Synovial fluid analysis.
Synovial fluid analysis in RF-negative polyarthritis reveals a nonspecific inflammatory reaction that is indistinguishable from characteristics found in other JIA subtypes. In children with polyarthritis (including those with extended oligoarthritis) polymorphonuclear neutrophils tend to be higher in synovial fluid compared with persistent oligoarticular disease but not significantly different from findings in systemic JIA. Cytokine, chemokine, and proteome profiles in synovial fluids have been explored in JIA but not sufficiently elucidated to be of clinical utility.
Radiographic evidence of joint space narrowing (decreased joint space, ankylosis, and carpal collapse) was demonstrated in 12% of 39 children at 2 years after onset and in 43% by 6 years after onset in the RF-negative JIA subset. Erosions and growth abnormalities likewise increased with time.
The limited information about synovial pathology indicates that the histological appearance of the synovium is similar for all JIA subtypes although there is greater vascularity in the polyarthritis group compared with enthesitis related arthritis, psoriatic arthritis, and oligoarthritis. Finnegan et al. reported knee joint synovial membrane histopathological and immunopathological features among 42 children with JIA of whom 8 had seronegative polyarthritis. All subjects were within 2 years of diagnosis (mean 5.3 months) and had not been treated with steroids or disease-modifying antirheumatic drugs (DMARDs). When compared with persistent oligoarticular JIA pathology, subjects with polyarthritis (RF positive, RF negative, and extended oligoarthritis) had more inflammatory infiltrates and synovial hypertrophy, and more abundant CD3, CD4, and CD20 cells. The adhesion receptor αVβ3 integrin, which is upregulated on endothelial cells during angiogenesis, was significantly more prominent in the polyarticular subjects than in either oligoarticular or extended oligoarticular patients.
As with all forms of chronic childhood arthritis, RF-negative polyarthritis requires a multifaceted approach to management. The mainstays of treatment include early and judicious use of pharmacotherapy, physical and occupational therapy, and the promotion of healthy lifestyles, including optimal nutrition, physical activity, and reduction of stress. Achieving and sustaining complete disease control is now an attainable objective.
An initial trial of nonsteroidal antiinflammatory drugs (NSAIDs) is appropriate. The trend to more aggressive treatment at first presentation of polyarthritis has resulted in the early use of a disease-remitting agent, usually methotrexate, often in combination with an NSAID; earlier initiation of biological therapies has also been advocated.
American College of Rheumatology (ACR) guidelines for treating JIA promote early treatment of JIA with rapid escalation of therapies as required in accord with responses. In RF-negative polyarthritis the aggressiveness of therapy can be guided by prognostic and disease activity indicators. Unfavorable prognostic indicators, which would prompt more aggressive early treatment, include hip and cervical spine involvement, radiographic evidence of joint space narrowing and/or bone erosions, and the presence of ACPA. Indicators of moderate to severe disease activities, judged by the number of active joints, levels of inflammatory markers, and poor physician and patient/parent global assessments would also prompt aggressive, early treatment.
There are no established guidelines for stopping drug therapy in JIA. Once disease remission is induced with a remittive agent, NSAIDs can often be discontinued without exacerbation of disease activity. In any event, even in polyarthritis with a favorable prognostic profile and low disease activity score, failure of NSAIDs to control the disease within 6 weeks should prompt the addition of methotrexate. Methotrexate is usually given by mouth initially, in doses of 10-15 mg/m 2 /week. In the absence of an adequate response the dose can be increased to 15-20 mg/ m 2 /week, preferably administered subcutaneously. The response to methotrexate is usually excellent. Nonresponse to methotrexate in RF-negative polyarthritis, as in other JIA subtypes, might be predicted by considering SNPs of genes involved in methotrexate metabolic pathways. For patients who are unresponsive to or intolerant of methotrexate, leflunomide is an option, although there is insufficient information to evaluate leflunomide’s role in RF-negative polyarthritis specifically. Treatment approaches that include use of biologically based anti-TNF agents at first presentation either in combination with a DMARD or alone have been suggested.
Anti-TNF agents are effective in treating children with polyarthritis who are unresponsive to methotrexate or leflunomide alone although there is little information to indicate that RF status correlates with responsiveness. Anti-TNF therapy should be considered in any child with polyarticular JIA with moderate or high disease activity who has not responded to methotrexate or leflunomide by 3 months or, for those with low disease activity, by 6 months. Biologically based therapies targeting proinflammatory cytokines other than TNF, including IL-6 as an example (tocilizumab), are emerging as treatment options. When patient-specific biomarker profiling becomes more clinically accessible, the selection of specific anticytokine therapies can be more biologically based.
Glucocorticoids are important as intraarticular therapy. Breit and colleagues reported a longer median duration of response to intraarticular triamcinolone hexacetonide in children with juvenile chronic arthritis who were RF negative (105 weeks) than in those who were RF positive (63 weeks). Glucocorticoids have a limited role as systemic therapy in polyarthritis although judicious use of systemic steroids as a bridging agent can be considered until disease-modifying agents become effective. Gold compounds and penicillamine are seldom used since the advent of generally safer and more efficacious pharmacotherapeutic options. Although hydroxychloroquine is at times used in RF-negative polyarthritis as an adjunctive and relatively safe agent, often in combination with methotrexate, there is no substantive evidence reported to support its efficacy.
There are no established guidelines to direct when pharmacotherapies can be safely discontinued in RF-negative polyarthritis. Although there is no evidence-based rationale for continuing treatment beyond the point of clinical remission, there is accumulating concern that clinical remission might not always reflect actual biologic remission. Both magnetic resonance and ultrasound imaging indicate that synovial inflammation can persist even in patients in whom clinical remission is achieved. Undetected, subclinical inflammation leads to joint damage and disability. Persistent biologic activity is evidenced by expression of TNF-α and IL-4-regulated genes and elevated levels of S100 proteins in patients in apparent clinical remission. In the future it is expected that incorporation of clinical, imaging, and biomarker indicators will be possible and will help identify true disease remission and thus guide the timing of when to safely stop treatment.
Exercise and Physical and Occupational Therapy
Regular participation in physical activity by children with JIA is beneficial. Functional impairment generally correlates with the extent and severity of articular disease, but poor fitness also occurs even in those with mild symptoms and persists even after disease remission. Both aerobic and anaerobic exercise capacity are decreased in children with polyarthritis compared with those with oligoarticular disease; those with RF-positive polyarthritis are somewhat more limited than those with RF-negative disease. Notwithstanding the advantages of active exercise, it is important to have a carefully designed passive therapy program. Children tend to function within the range of motion they have, not the range they should be trying to achieve. Focused physical therapy should be instituted as soon as inflammation subsides sufficiently to facilitate the child’s cooperation. Physical therapy aimed at restoration of range of motion can be facilitated by pretreatment with an analgesic or application of heat. Major contractures are often more amenable to therapy after intraarticular triamcinolone hexacetonide (see Chapter 14 ).
The need for surgical management is now less common as a consequence of more effective medical management. Nonetheless, some children with resistant or untreated disease will require joint replacement of hips, knees, or, less commonly, other joints. Prior to surgical procedures the child with polyarthritis should be evaluated for conditions that might present added anesthesia and surgery risk, including cervical spine and TMJ damage, immunosuppression that heightens infection risk, and poor bone quality that can compromise the integrity of joint implants.
Course of the Disease and Prognosis
RF-negative polyarticular JIA is a chronic disease, lasting years or decades. Oen et al. reported that only 25% of 80 children with RF-negative polyarthritis diagnosed between 1977 and 1994 and followed for at least 5 years had gone into remission by the age of 16. Children who had not remitted by this age were likely to have ongoing active arthritis into their late 20s or early 30s. These earlier data indicate that RF-negative polyarticular JIA was associated with substantial morbidity in most affected children. More recent studies, reflecting advances in treatment, suggest more favorable short-term outcomes. However, based on results of a prospective study, Oen et al. reported that only 19% of RF-negative polyarthritis patients achieved clinical remission within 6 months of enrollment. One third to half of patients with a severe polyarticular disease course who fail treatment with methotrexate achieve ACR Pedi 70 criteria response to an anti-TNF agent within 12 months. Greenwald et al. studied an older onset subset of patients with polyarticular JIA and found almost 50% had a favorable functional outcome at 2 years even before the era of biological therapies. By 6 months after initial diagnosis, almost half of the patients had fewer than five active joints and by 24 months all patients had fewer than five active joints, regardless of how many active joints they had at diagnosis. More favorable medium- and long-term outcomes are anticipated in the RF-negative polyarthritis group as a result of treatment advances.
Most children with RF-negative polyarthritis will retain their classification category; however, manifestations of psoriasis or inflammatory bowel disease, as examples, might first manifest months or years following the onset of arthritis, thereby necessitating reclassification.
Rheumatoid Factor Positive Polyarthritis
RF-positive polyarticular JIA shares a similar clinical phenotype, serology, and immunogenetic profile to that of adult rheumatoid arthritis, and both can occur in the same family. Because of these features, RF-positive polyarthritis is generally accepted to be equivalent to RF-positive adult rheumatoid arthritis (RA) but with onset at younger than 16 years of age. The association with ACPA emphasizes this relationship.
Incidence and Prevalence
There is limited published information about incidence and prevalence rates for the RF-positive polyarthritis onset subtype. Estimated incidence rates of 0.3 to 0.7 per 100,000 person-years at risk have been reported or can be calculated from publications from Europe and the United States. In comparison, incidence estimates as high as 12.3 can be calculated from published data for East Coast Alaskan Indian children, and 8.1 per 100,000 person-years at risk for Native Canadian Indian children in Manitoba. Similarly, estimates of point prevalence of 0 to 6.7 per 100,000 in Europe and the United States, and 54 per 100,000 at risk for Manitoba Canadian Indian children, can be calculated from published data.
Age at Onset and Sex Ratio
The mean age at juvenile onset of RF-positive polyarthritis is 9 to 11 years, and the range is 1.5 to 15 years. Affected girls outnumber boys from 4 : 1 to 13 : 1 in large series.
Geographic and Racial Distribution
RF-positive polyarthritis is one of the least frequent JIA subtypes in white children but occurs at higher relative frequencies in non-white children. Reported frequencies range from 51% in a series of Native Canadian Indian children to 18% of Hispanic-American children, 7% to 17% of East Indian children, 14% of African-American children, and 12.5% of Japanese children, and 0.2% to 6% in European and US children and adolescents with chronic arthritis.
Etiology and Pathogenesis
ACPA occur in 60% or more of patients with RF-positive polyarthritis, and may have a role in disease pathogenesis. ACPA are directed against citrullinated peptides that are formed in dying cells by deimination of arginine by peptidylarginine deiminases. It is postulated that these modified peptides are exposed to the immune system if clearance does not occur. Moreover, peptidylarginine deiminases leak from dying cells, allowing citrullination of peptides in tissues such as the synovium. It has been shown that shared epitope–bearing HLA-DR alleles have increased binding affinity for citrulline. This phenomenon is a most likely explanation for the association of the SE specifically with ACPA-positive JIA and RA. Further, a pathogenic role of ACPA is suggested by their presence prior to clinically evident RA, the increased severity of erosive disease in ACPA-positive patients with RA or JIA, and the induction of arthritis by ACPA in experimental animals. ACPA immune complexes can activate complement in vitro, and therefore complement activation through complexes of ACPA and their tissue targets may be a mechanism of disease. These postulates likely apply to ACPA-positive RF-positive polyarticular JIA as much as they do to ACPA-positive RA.
Although disease initiation may differentiate RF-positive polyarthritis, it shares inflammatory pathways with other JIA subtypes. These include chemokine and cytokine secretion, angiogenesis, the infiltration of inflammatory cells into the synovium, synovial cell hyperplasia, and finally secretion of matrix metalloproteinases (see Chapters 3 and 4 ).
Distinct from RF-negative polyarticular JIA, RF-positive polyarticular JIA shares genetic predispositions with adult RA. The shared epitope (SE), a specific sequence present on a number of HLA-DRB1 antigens, is associated with increased risk of both RF-positive polyarthritis and adult RA. The SE is found on HLA-DRB1*04 (0401, 0404, 0408, or 0405), DRB1*01 (0101), and DRB1*14 (1402) alleles. Population frequencies of the SE and particular SE-bearing HLA alleles vary in different ethnic groups. Thus, RF-positive polyarthritis and RA are associated with DRB1*04 alleles, mainly DRB1*0401 and *0404, in white populations. The relative risk of RF-positive polyarthritis attributable to DRB1*0401 is 3.2 to 7.2, and to DRB1*0404, 3.8 to 8.9. Double doses of the SE further increase the relative risk of the disease. The associated allele is DRB1*0405 in Asian (i.e., Japanese, Chinese, Korean) populations. In Native North American Indian tribes the SE is represented by DRB1*04 alleles and DRB1*1402 and is associated with RA. In some tribes the population frequency of DRB1*1402 is so high that no significant increase is found in patients with RA. In Native Canadian Indian children, the situation is more complex when both the SE and DRB1*0901 occurring together as a genotype are associated with RF-positive polyarthritis. This dual association supports the suggestion that earlier age at onset is associated with a greater genetic influence.
Population frequencies of the SE tend to correlate with frequencies of RA and RF-positive polyarticular JIA. For example, the frequency of the SE in white populations is 27% to 36%, whereas Native North American Indian populations with high incidence and prevalence rates of RA and RF-positive polyarthritis have frequencies of 66% to 98%. As discussed above, the SE is associated specifically with ACPA-positive, rather than RF-positive, disease.
HLA-DR alone accounts for 17 % of the genetic risk for juvenile rheumatoid arthritis (JRA) and provides the greatest risk with odds ratios (OR) ranging from 3 to 12. Moreover, HLA associations are specific for JRA or JIA subtypes. In contrast, although the remaining genetic risk is large, it is contributed by multiple genes, each providing only modest probabilities with OR values less than 1.5, with the exception of PTPN22. Many associations with non-HLA genes are shared by a number of autoimmune conditions, suggesting common immunopathogenic pathways. The associated genes are often called autoimmune genes and many are involved in T-cell activation and differentiation. Autoimmune genes with replicated associations with JIA include SNPs of TNF-α–induced protein 3 (TNF AIP3), STAT-4, IL-2RA, PTPN22, and PTPN2 genes, whereas associations with TRAF1/C5, IL2RA gene variations are controversial. Not all studies have included patients with RF-positive polyarthritis, and few have stratified patients by JIA subtype. One study showed a suggestive association of RF-positive polyarthritis with a STAT4 SNP and another with a SNP in the MVK gene, the gene for hyper-IgD syndrome. However, as with HLA, it is likely non-HLA gene associations for RA are shared with RF-positive polyarthritis. These associations include all the autoimmune genes mentioned above. In contrast to shared autoimmune genes, the peptidylarginine deiminase 4 (PADI4) gene association is specific for RA, particularly for ACPA-positive RA, but this association is strongest in East Asian populations. The PADI4 genes code for PAD enzymes. There is a possible functional correlation as the associated PADI4 haplotype stabilizes PADI4 messenger RNA (mRNA), allowing increased citrullination. An association of PADI4 polymorphisms specifically with RF-positive polyarthritis has not been investigated.
Upper and lower extremity large and small joints are affected, as well as the cervical spine and TMJs. The thoracic and lumbar spine and sacroiliac joints are spared. Although large joints are commonly involved, the characteristic pattern is symmetrical arthritis affecting the metacarpophalangeal (MCP) and proximal interphalangeal (PIP) joints of the hands, the wrists, and the metatarsophalangeal (MTP) and PIP joints of the feet. In contrast to RF-negative polyarthritis, micrognathia (TMJ involvement) does not usually occur because of the individual’s older age at onset. Early limited range of motion occurs at the wrists and can eventually progress to more substantial debility and deformity. Deformities that develop at the hands include ulnar drift at the wrists and the MCP joints, and boutonnière and swan neck deformities at the fingers ( Fig. 17-5 ). Deformities that develop at the feet include hallux valgus deformity at the first MTP joints, hammertoe, and cock-up toe deformities.
Fatigue and weight loss may occur with active disease. Fatigue may persist even if disease is inactive. Fever is rare in RF-positive polyarthritis, and a rash does not occur.
Other than nodules, uveitis or extraarticular disease manifestations associated with adult RA (described below) rarely occur in patients with RF-positive polyarthritis, whether during childhood, adolescence, or adulthood.
Frequencies of uveitis of 1.4 and 4.5% in children with RF-positive polyarthritis have been reported in large series.
The most common extraarticular signs in patients with RF-positive polyarthritis are rheumatoid nodules. In Ansell’s series, 30% of patients with polyarticular RF-positive arthritis had rheumatoid nodules during the first year of disease. Nodules often occur distal to the olecranon and at other bony prominences and pressure points, on flexor tendon sheaths, Achilles tendon, and on the soles of the feet. They are firm, mobile, and nontender; however, pressure of the nodule against soft tissues or bone may cause pain. The presence of rheumatoid nodules indicates a poor prognosis. Accelerated nodulosis may occur in patients on methotrexate. In this case the nodules are multiple, develop over a short time, tend to occur on the hands, and regress on discontinuation of methotrexate. This complication has been described in two children with RF-positive polyarthritis and four with systemic JIA. Methotrexate-induced nodulosis is associated with minimal discomfort and may stabilize with use of hydroxychloroquine. Nodulosis associated with methotrexate does not necessarily preclude continuation of methotrexate therapy.
Rheumatoid nodules must be distinguished from subcutaneous nodules of rheumatic fever, which are smaller; so-called benign rheumatoid nodules that are not associated with chronic arthritis; granuloma annulare, which are small nodules arranged in a circular pattern; erythema nodosum; and nodules seen in cutaneous polyarteritis nodosa.
Rheumatoid vasculitis is rarely described in RF-positive polyarticular JIA during childhood or adolescence. In 1978, Ansell noted nailfold and extensive cutaneous vasculitis in several patients during prolonged follow-up. However, the lack of reports of this complication in recent literature may reflect improved therapies for arthritis or less severe disease, as vasculitis in adults with RA tends to occur in those with the most severe disease.
Felty syndrome consists of persistent neutropenia, splenomegaly, and RA, and is associated with frequent infections. The bone marrow is normocellular, and the mechanism of neutropenia is complex, involving both antigranulocyte antibodies and decreased granulopoiesis. In adults, Felty syndrome occurs in RF-positive patients with long disease duration. It has been reported rarely in adolescents with RF-positive polyarthritis and in adults who had juvenile onset disease.
Cardiovascular and pulmonary disease.
Valvular heart disease has been reported in at least eight patients with childhood-onset RF-positive polyarthritis. Aortic insufficiency is the most common lesion. Patients present with sudden onset of congestive heart failure or may deteriorate suddenly after a variable period of stability following the detection of cardiac murmurs. Valve replacement is almost always required. Cardiac symptoms may start during childhood, adolescence, or adulthood, at intervals varying from 4 to 17 years from onset of JIA. However, pathological murmurs may be detected as early as 1 year after onset. Patients with JIA who have organic cardiac murmurs should be evaluated for valvular insufficiency and monitored carefully.
Pulmonary parenchymal disease, so-called rheumatoid lung, has been reported in seven cases of RF-positive polyarthritis. Two types of pulmonary involvement have been reported: lymphoid interstitial pneumonitis, and bronchiolitis obliterans or bronchiolitis obliterans organizing pneumonia (BOOP). These pulmonary complications may occur during childhood and adolescence or in adulthood. The time interval between the clinical presentation of pulmonary disease and onset of JIA has ranged from 10 years before to 20 years after onset of JIA. Symptoms include tachypnea, dyspnea, a nonproductive cough, and fever. On auscultation, crackles and an end-inspiratory squeak are often heard. Diagnosis is based on clinical history and findings, pulmonary function tests, chest radiographs, and high-resolution computed tomography (HRCT). Bronchoalveolar lavage and/or lung biopsy may be necessary. Pulmonary function tests show reduced lung volumes and decreased diffusion capacity. A restrictive pattern is seen when interstitial pneumonitis is present, and an obstructive pattern is seen in BOOP. Chest radiographs may be normal or may show interstitial infiltrates. HRCT abnormalities include ground glass changes suggesting inflammation, bronchiectasis, or bronchiolectasis (suggesting BOOP), and honeycombing (suggesting fibrosis). The differential diagnosis includes drug-induced pulmonary toxicity and infection. The prognosis of rheumatoid lung is variable in children and adolescents. Although a few patients have improved with corticosteroid therapy, others have deteriorated despite corticosteroid and immunosuppressive therapy.
The differential diagnosis of polyarthritis is discussed above. Specific diagnoses to be considered in the context of RF-positive polyarthritis are connective tissue diseases, reactive arthritis, and infections, in which polyarthritis and a positive test for RF may occur concurrently. Among the connective tissue diseases, SLE and overlap syndromes, including mixed connective tissue disease, are diagnostic considerations in the child or adolescent with polyarticular arthritis who has a positive test for RF. RF is positive in 10% to 30% of children with SLE and in approximately 66% of children with mixed connective tissue disease. RF may be present in cases of acute rheumatic fever. Tuberculosis and subacute bacterial endocarditis can be associated with arthritis accompanied by a positive test for RF.
Indicators of Inflammation
These are identical to those in RF-negative polyarthritis discussed earlier.
The classification of RF-positive polyarthritis is based on the presence of two positive tests for RF performed at least 3 months apart during the first 6 months of disease. Patients with RF-positive polyarthritis are characterized by persistently positive IgM RF, generally in high titer.
Approximately 42% to 56% of children with RF-positive polyarthritis have antinuclear antibodies.
Anticitrullinated protein antibodies (ACPA).
ACPA are much more prevalent in RF-positive polyarthritis than in other JIA subtypes; however, as in adults with RA, the concordance with RF positivity is not complete. The frequency of ACPA in RF-positive polyarthritis varies from 57% to 90% (mean 74%); but up to 17% of children with RF-negative polyarthritis, and 6% of children with subtypes other than RF-positive polyarthritis, have positive ACPA tests. ACPA correlate with disease severity and joint damage evidenced by radiographs, suggesting that ACPA have a prognostic significance in JIA as in RA.
Synovial fluid analysis.
Synovial fluid analyses from patients with RF-positive polyarthritis show an inflammatory fluid not clearly differentiated from that found in other forms of JIA. Synovial fluid cell counts and proportions of neutrophils may be higher in RF-positive polyarthritis than in those with oligoarticular arthritis.
Most information on joint damage in RF-positive polyarthritis comes from a limited number of studies of plain radiographs of patients treated before the introduction of biological therapies. Joint space narrowing and erosions occur within the first 1 to 2 years after onset and are most frequent at the wrists, hands, feet, and shoulders ( Fig. 17-6 ). At the wrist, cartilage loss occurs at the proximal wrist joint and in the intercarpal joints, resulting in carpal ankylosis and shortening. Both erosions and cartilage loss occur more frequently in RF-positive polyarthritis than in other forms of JIA. Atlantoaxial subluxation of the cervical spine is more frequent in patients with RF (36% frequency) than in other patients with JRA (16%). In patients with long disease duration extending into adulthood, radiographic damage at the hands and feet is more frequent in RF-positive (91%) than RF-negative polyarthritis (55%), but hip damage is more frequent in RF-negative polyarthritis (20% vs 50%); the frequency of all radiographic lesions at the cervical spine is similar, at 65%.