Psoriatic Juvenile Idiopathic Arthritis

Incidence and Prevalence

The incidence and prevalence of PsJIA are unknown. Population data, enumerating largely patients with adult-onset PsA, suggest a prevalence of 0.10% to 0.25% in the United States. ^, It can occur in all ethnic groups. The proportion of JIA patients with PsJIA varies widely depending on the population studied and the diagnostic criteria employed. Series that recognize patients on the basis of frank psoriasis, or using ILAR criteria, find that PsJIA represents approximately 7% (range: 0% to 11.3%) of patients with JIA. ^{, , , , ,}

Age at Onset and Sex Ratio

In the pediatric population, the age at onset of PsJIA is bimodally distributed ( Fig. 21.1 ). ^, A first peak occurs during the preschool years, and a second is seen during middle to late childhood. As with most forms of JIA, PsJIA is very uncommon before the age of 1 year. Girls account for 60% of patients in larger series, including approximately 75% of those with early-onset disease versus closer to 50% of those with late-onset disease ( Table 21.2 ). ^,

Synovial Pathology

Pathological data concerning the psoriatic synovium are available largely from adult-onset disease, with rare exceptions. ^, Gross examination, as performed by arthroscopy, reveals a synovial lining that is less villous than in adult RA but with distinctive tortuous, bushy superficial blood vessels. ^, This microvascular pattern resembles that of the psoriatic plaque, and is observed also in synovial tissue from the spondyloarthropathies. ^, There are histological changes throughout the psoriatic synovium ( Fig. 21.2 ). The lining becomes hypertrophic with expansion of both synovial macrophages and synovial fibroblasts. The infiltrate in the loose connective tissue beneath the synovial lining is composed principally of lymphocytes and monocyte/macrophage lineage cells, with occasional neutrophils, plasma cells, and mast cells. Lymphoid follicles may be observed. Recently a population of myofibroblast-like cells has been identified in synovium from both psoriatic and nonpsoriatic spondyloarthritis, but the identity and significance of these cells remains unknown. Compared with RA, lining hypertrophy and sublining infiltrates are typically less extensive. Infiltrating neutrophils are more prevalent in PsA, but they are not invariably present. ^{, ,} In general, given the variability between patients and within different parts of the same synovium, pathological findings are inadequate to define the diagnosis in an individual patient, and biopsy is rarely performed as a diagnostic tool in PsJIA.

Characterization of the psoriatic synovial infiltrate by immunohistochemistry shows that the majority of infiltrating lymphocytes are T cells that express the memory CD45RO phenotype, with CD4 helper cells predominating over CD8 cytotoxic cells. ^{, ,} These cells are present at frequencies similar to that in RA, as are CD20 ⁺ B cells, plasma cells, and CD68 ⁺ macrophages. T-cell oligoclonality suggests local antigen-driven expansion. CD83 ⁺ dendritic cells are less common than in RA, whereas interleukin (IL)-17–expressing mast cells are more common. Increased numbers of macrophages expressing CD163 are identified in adult PsA and the adult spondyloarthropathies, ^, though not as clearly in juvenile-onset disease. CD163 is a scavenger receptor, typically expressed on mature resident tissue macrophages that may help limit rather than promote inflammation. However, the activity of these cells in the psoriatic synovium is unknown. ^,

Complement-fixing immune complexes are not typically found in the psoriatic synovium, and synovial fluid complement levels are usually normal. Similarly, citrullinated peptides are observed commonly in the rheumatoid synovium but rarely in PsA.

Entheseal Pathology

Work by Dennis McGonagle and others has identified entheseal pathology as a major feature of spondyloarthropathies, including PsA, as well as a characteristic difference between PsA and RA. Evidence to support this hypothesis is derived from histological, imaging, animal, and even human epidemiological data.

Entheseal sites are not readily accessible to biopsy, but small series in adult spondyloarthropathies provide a degree of histological insight. A low-grade inflammatory infiltrate is observed, often in association with underlying erosion of bone. This infiltrate is not limited to the surface of the bone, and it is often more extensive in the bone marrow underlying the enthesis. Cells observed at the interface include macrophages, lymphocytes (particularly CD8 ⁺ T cells), and occasional neutrophils. A case report of a child with psoriatic dactylitis who underwent biopsy to evaluate for infectious causes identified a strong CD4 ⁺ T-cell infiltrate in the tenosynovium. Consistent with the synovial data described earlier, this was noted to be hypervascular as well. Bone healing is commonly evident, with woven bone filling in the defect left by erosions. This new bone often extends beyond the previous bony surface to interface with the ligament. These observations have given rise to the hypothesis that the new bone formation characteristic of the spondyloarthropathies results from recurrent cycles of injury and healing, perhaps enabled by fluctuations in the degree of inflammation. ^, Whether such a mechanism underlies the hypertrophic periostitis observed in some patients with PsJIA is unknown.

Multiple studies have used advanced imaging modalities to evaluate the location of pathology in PsA, often with RA as a disease control. Whereas RA small joint inflammation is typically intraarticular inflammation accompanied perhaps by tenosynovitis, substantially more extraarticular inflammation is observed in PsA, including osteitis, tenosynovitis, flexor and extensor tendon enthesopathy, and soft tissue edema. ^, Imaging of fingers with dactylitis demonstrates flexor and extensor tendon edema, soft tissue edema, and further edema specific to the so-called pulleys , which are the regions where the tendons wrap around the bones. Such pulleys are present not only in the small joints of the fingers, but also around the ankles, where the tibialis posterior wraps around the medial malleoli en route to the foot. An ultrasound study showed that the pulleys were thicker in PsA compared with RA despite the extensive finger involvement characterized by the latter condition. There is also an association between psoriatic nail disease and associated extensor tendon enthesopathy. ^, The term synovio-entheseal complex was created to describe the strong link between entheseal and articular pathology present in PsA.

The concept of an entheseal basis to SpA is supported as well by animal work. In the model characterized by tumor necrosis factor (TNF) overexpression, sequential histological studies revealed that inflammation begins in the synovio-entheseal complex, spreading to the neighboring joints by age 6 to 8 weeks. Likewise, in the model associated with IL-23 overexpression, IL-23R ⁺ T cells reside in the entheses and respond to systemic IL-23, inducing systemic and local inflammation.

It has long been suspected that SpA is triggered at least in part by biomechanical forces. Indeed, in the TNF overexpression model, disease is abrogated when animals are made non–weight-bearing. A large body of epidemiological data in humans supports this hypothesis. For example, a longitudinal cohort obtained from the Health Improvement Network database containing millions of patients followed by general practitioners in the United Kingdom showed that among patients with psoriasis, those with diagnostic codes suggestive of trauma were at increased risk of subsequent development of PsA, particularly if the trauma involved the bone or joint. In the same cohort, trauma did not appear to increase the risk of RA. Additionally, in patients with PsA, occupations involving repetitive stress were associated with increased risk of radiographic progression. Even heavy smartphone use is associated with disease activity in PsA, an issue clearly relevant to the pediatric population. These findings are said to be reflective of the so-called deep Koebner phenomenon, named after the observation of psoriatic flares triggered by skin trauma.

Genetic Contribution

There is convincing clinical evidence of a genetic contribution to psoriasis and PsA. More than 50% of patients with childhood-onset psoriasis, with or without PsJIA, have a family history of psoriasis ( Table 21.2 ). ^{, ,} The risk for both psoriasis and PsA appears to be transmitted more effectively via the paternal line (genetic imprinting). ^, A 50-fold increased risk of PsA was observed in family members of adults with PsA, suggesting that a propensity for arthritis is inherited over and above the propensity for psoriasis. Similar results were noted in other studies. ^,

Association studies have begun to shed light on the genes that explain these strong familial associations. ^, In adults, psoriasis with onset age 40 years or younger (type I psoriasis) is more strongly familial than older-onset (type II) disease. ^, Type I psoriasis is strongly associated with the major histocompatibility complex (MHC) class I allele HLA-Cw6. This allele is also associated with adult PsA, and possibly with older-onset PsJIA in children, but the link appears secondary to risk for psoriasis. ^, Genome-wide association studies in patients with PsA have identified a clear association with HLA, as well as associations with several other genes observed in other forms of arthritis, including IL-23R and IL-12. Certain HLA alleles, including HLA-B27, are overrepresented in adult patients with PsA compared with psoriasis controls. Additionally, genotype/phenotype studies demonstrate that specific alleles are associated with their own associated clinical findings. For example, HLA-B27 is associated with a shorter interval from onset of psoriasis to development of arthritis, as well as with symmetrical sacroiliitis, whereas HLA-B∗0801 is associated with asymmetrical sacroiliitis yet more severe synovial pathology, enthesitis, and dactylitis. In contrast, HLA-C∗0602 is associated with psoriasis more so than PsA and may even be protective against enthesitis. ^, The results of studies of HLA associations in PsJIA have been inconsistent, likely because of differences in definitions employed and variability within PsJIA across the pediatric age spectrum. ^{, , , , ,} However, there do appear to be different alleles in children with older versus younger age of onset. ^,

Beyond HLA, PsJIA has been linked with single nucleotide polymorphisms (SNPs) near genes involved in the autoinflammatory diseases ( MEFV, NLRP3, NOD2, and PSTPIP1 ). These associations have not emerged in adult genome-wide association scans and have yet to be replicated. In adult studies, as noted earlier, psoriasis and PsA have been associated with SNPs in a range of genes, including HCP5 (involved with control of viral replication) and genes related to the cytokines TNF, IL-12, IL-13, and IL-23. ^, The association with IL-23R has been corroborated in PsJIA. The functional consequences of these SNPs remain to be determined, but the cytokine findings are of particular interest in light of established and emerging therapeutic targets (see section Treatment). IL-23 is involved in the differentiation of proinflammatory T helper (Th)17 cells, which are increased in frequency in the circulation and joints of patients with PsA and are present in psoriatic plaques. IL-23–responsive cells are also found in entheses and the aortic root, and IL-23 overexpression in mice can induce a spondyloarthritis phenotype with inflammation at both sites. IL-13 suppresses the Th17 axis in favor of differentiation along a Th2 pathway, and the “risk” allele linked to psoriasis is associated with decreased cytokine production. Murine models suggest that Th17 cells may contribute importantly to arthritis. The implication of these findings is that the Th17 axis may be important in both psoriasis and its associated arthritis.