The Juvenile Psoriatic Arthritis Cohort in the CARRA Registry: Clinical Characteristics, Classification, and Outcomes

DISCUSSION

The CARRA-JPsA cohort studied here represents 5.6% of the JIA population in the CARRA legacy registry, in line with published data for prevalence of this subtype within JIA^5,12,13,14. Data on this cohort were analyzed in light of current knowledge and understanding of PsA from both pediatric and adult rheumatology literature using the ILAR and CASPAR classification criteria.

A bimodal age of onset has been reported in children with JPsA^4,14,15, similar to the well-recognized yet poorly understood demographic feature of all JIA, which has 1 peak age of onset at 2–3 years and a second peak in mid-adolescence¹⁶. Our findings are consistent with the literature pointing to a female predominance in early-onset JPsA, and increased incidence of sacroiliitis in later-onset JPsA. Stoll and Punaro observed that patients with early-onset JPsA had features similar to ANA-positive oligoarticular and polyarticular JIA, including more females and presence of chronic uveitis, whereas older children had features of SpA¹⁴. The better outcome observed in females with early arthritis onset may be explained in part by the inclusion of females with oligoarticular arthritis, who in general have a more mild disease¹⁷.

The influence of age and sex on disease expression, which is also observed in adults with PsA, likely has a genetic basis^18,19,20. Adult patients with early-onset PsA (onset age < 40 yrs) typically have a longer psoriasis-arthritis latency period, a family history of disease, severe psoriasis, enthesitis, and oligoarthritis²¹. Additional studies point to more frequent axial involvement in males¹⁸, whereas presence of polyarthritis correlates positively with age and female sex^19,22. Taken together with these results, our data suggest an overall pattern in which JPsA presents in childhood (at ≤ 4–6 yrs) as predominantly a disease of females with peripheral arthritis, followed by presentation of later-onset JPsA (> 4–6 yrs) and early PsA with enthesial and axial involvement, more common in males, and then a later presentation again with more tendency for peripheral arthritis. The precise inflection point varies by the study and is subject to errors in symptom recall. One interpretation is that later-onset JPsA is an early presentation of early-onset adult PsA disease. It is also possible that early-onset “JPsA” is a subset of early-onset (nonsystemic) JIA as a whole, rather than a discrete subset of JPsA. Further studies of the JPsA population are needed for a better understanding of the influence of age and sex on the clinical manifestations of the disease. It will be interesting to compare data from the young onset subgroup with the like-aged patients with JIA, in the CARRA registry and in other datasets.

CARRA-JPsA patients in our cohort tended to have less psoriasis compared with adult patients, in whom psoriasis will precede or be diagnosed at the time of arthritis in 85% of patients²³. This observation was made in other JPsA cohorts as well^4,5. A possible explanation is that, in contrast to adults with PsA, psoriasis in JPsA often occurs years after arthritis onset^5,24,25, and its development may be obscured by DMARD treatment, as has been suggested for uveitis associated with JIA²⁶. Only 37% in our CARRA-JPsA cohort report nail involvement, compared with 57% in previous JPsA studies^2,5,15 or 66% in adults²⁷. The prevalence of dactylitis in our cohort (29.7%) is similar to the report of Butbul Aviel, et al², but lower compared with older reports^5,15. The prevalence of enthesitis and uveitis in our cohort are in the range (14%–45%) of published series^2,4.

Axial involvement, documented as sacroiliitis/IBP in the CARRA-JPsA cohort, was found in 16.7%, and axial changes were found in 21.6% of those who had radiographic imaging. Pediatric and adult literature point to asymptomatic axial involvement in patients with psoriasis and to axial radiographic features of SpA^15,28,29. There was no correlation between HLA-B27 positivity and radiographic or clinical findings of sacroiliitis, in line with a report on PsA³⁰.

At enrollment, an average 4.6 years after symptom onset, 24.6% of patients in the CARRA-JPsA cohort had radiographic evidence of joint damage, similar to findings reported by Southwood, et al in 1989⁵. However, taking into consideration the bDMARD prescribed to 52.9% of our cohort, these results are a cause for concern and point to the aggressive nature of the disease. Supporting this idea, Kane, et al described erosive changes after 2 years of followup in 47% of patients with early PsA, despite treatment with DMARD³¹. The use of DMARD in our CARRA-JPsA cohort is in line with that reported in the registry as a whole³².

The prevalence of ANA positivity in our cohort is similar to that found in other JPsA studies^2,4,5,15 and series in PsA^33,34. RF positivity (4.7%) in our cohort is also found in PsA³⁵, although ACPA positivity is lower, 1.9% vs 10.6%^36,37. Differences in laboratory methods may have confounded these results.

Our findings documenting improvement in objective findings are supported by Butbul Aviel, et al, whose study was also performed in the era of bDMARD and found the median time to inactive disease while taking therapy was 1.23 years². To our knowledge, only 1 study to date has addressed longterm prognosis in JPsA. Of 31 patients followed for ≥ 15 years, 55% were in remission without medication for ≥ 12 months; however, 33% had active disease requiring DMARD treatment³. Our observation of a favorable prognosis in females with younger age at disease onset is not supported by others^2,4. Possible explanations include the short (1-yr) followup in our study and differences in study populations^2,4. The potential of PsA to cause impaired well-being and function beyond objective variables³⁸ was demonstrated in our CARRA-JPsA cohort. This finding is similar to findings in PsA in which a high prevalence of fibromyalgia was observed³⁹. In our dataset at followup, objective improvements were not accompanied by improvements in HQ scores, although this may be in part because HQ scores consistent with overall good status already were registered at baseline.

The 19.5% of our CARRA-JPsA cohort whose disease manifestations did not fulfill ILAR criteria tended to have fewer classical manifestations of psoriasis, dactylitis, and nail pitting. Because psoriatic skin rash may manifest years after arthritis onset in children⁵, these findings may reflect the physician’s suspicion that PsA was the appropriate diagnosis, based on the presence of SpA and consequently, less strict application of ILAR criteria, by which those children would be assigned to ERA or undifferentiated JIA⁴⁰. The CASPAR classification criteria for PsA⁷ are likely to be useful as diagnostic criteria⁴¹. Applying those criteria to the CARRA legacy registry, the JIA cohort identified 52 additional patients who would be classified as PsA by adult rheumatologists. Those patients had less psoriasis but more enthesitis, IBP, IBD, and uveitis, all manifestations of the spondyloarthritides⁴⁰. Findings of asymptomatic enthesitis in patients with JPsA and PsA provide additional support for this concept^42,43.

Although some differences we report may reflect semantic differences in disease classification, they have therapeutic implications. Because certain traditional DMARD appear to be less effective in PsA compared with RA, whereas newer biologics — particularly those targeting interleukin 17 — appear to be more effective in psoriatic disease⁴⁴, patients who are not classified as having JPsA by pediatric rheumatologists — versus those labeled as having JPsA/PsA — might be offered different treatment options as new medications become available. Psoriasis, which often occurs after arthritis in children and may also be prevented by use of DMARD, does not serve as a reliable indicator. Further research on the immunology, pathogenesis, and genetics of these clinically related conditions may provide the biologic information needed for a more precise classification scheme⁴⁵.

Several limitations of our study should be noted. Our analyses were retrospective using classification criteria. Information on anterior versus posterior uveitis was not available, and articular involvement was recorded as oligoarthritis and polyarthritis, not number of involved joints. Information on the extent of arthritis during 1 year of disease, but not at onset, was collected, limiting our ability to compare JPsA subgroups defined by oligoarticular and polyarticular disease. Radiographic data were not interpreted centrally, and lacked specific information, such as the type of imaging performed. Thus, our algorithm assigned points for CASPAR criteria without the radiographic component; notably, however, this criterion is the least frequently used when the CASPAR criteria are applied⁷. In the implementation of the ILAR criteria we excluded children having uveitis, because the CARRA registry does not distinguish between acute and chronic uveitis. Therefore, underestimation of the ILAR-JPsA and CASPAR-JPsA cohorts could have occurred.

Our analysis of data on the physician-diagnosed CARRA-JPsA cohort in the CARRA legacy registry supports the division of this condition into 2 subgroups, according to the age of onset. After 1 year of followup, a significant improvement was noted in objective findings, but without improvement in HQ scores. We suggest that pediatric and adult rheumatologists should be familiar with the differences in their approaches; adult rheumatologists should be aware that they may be inheriting a number of patients with PsA (per CASPAR criteria) who might be labeled differently by their pediatric colleagues. Applying CASPAR criteria across the age spectrum would likely enhance continuity of care and better reflect current concepts about PsA pathogenesis and expression. This suggestion requires discussion and validation by a panel of experts in pediatric and adult rheumatology.