Long-term outcomes and disease course of children with juvenile idiopathic arthritis in the ReACCh-Out cohort: a two-centre experience

Abstract

Objective

To assess long-term outcomes of children with JIA diagnosed in the biologic era.

Methods

Chart review of patients prospectively enrolled in the Research in Arthritis in Canadian Children Emphasizing Outcomes inception cohort at two Canadian centres. Inactive disease and remission were defined according to Wallace criteria.

Results

We included 247 of 254 (97%) eligible patients diagnosed 2005–10. At the last follow-up visit at a median age of 16.9 years, 47% were in remission off medications, 25% in remission on medications and 27% had active disease; 51% were on at least one anti-rheumatic medication (22% on biologics). Patients with systemic JIA had the highest frequency of remission off medications (70%) and patients with RF-positive polyarthritis had the lowest (18%) (P <0.05 by Fisher’s exact test). Among 99 patients with oligoarthritis at enrolment, 14 (14%) had an oligoarthritis extended course. Forty-five patients (18%) had at least one erosion or joint space narrowing in X-rays or MRI, and two (0.8%) required joint replacement.

Conclusion

Relative to historical cohorts, this study suggests a reduction in JIA permanent damage, a more favourable prognosis for systemic JIA and a lower progression to oligoarthritis extended category. However, in an era of biologic therapy, one in four patients with JIA still enter adulthood with active disease and one in two still on treatment.

Rheumatology key messages

• Canadian children with arthritis diagnosed in the biologic era were assessed in early adulthood.

• Relative to historical cohorts, remission frequency increased (72%) and permanent juvenile arthritis damage decreased.

• One in four patients with juvenile arthritis had active disease and one in two were still on treatment.

Introduction

In the current era of evolving treatment strategies, up-to-date knowledge about long-term outcomes of patients with JIA is essential for patient counselling and planning transition to adult care. Outcomes may have improved since the pre-MTX and pre-biologic eras when many young adults had ongoing disease activity, deformities and disability [1, 2]. Among patients diagnosed 1974–94 at three Canadian centres, 41% had active disease at a median age of 18.8 years, and joint replacement was required in 17% of patients with systemic JIA and 23% with RF-positive polyarthritis [3]. By age 16 years, no patients with RF-positive polyarthritis, and only 36% of patients with systemic JIA were in remission [3]. Current practice emphasizes early diagnosis and individualized treat-to-target, often with MTX and biologics, aimed at achieving inactive disease [4]. Outcomes vary depending on JIA category and the presence of features of poor prognosis [5]. However, differences in disease classification criteria, available therapies, study design and duration of follow-up in published cohorts limit their application for counselling families of children newly diagnosed with JIA in Canada. We sought to assess the long-term outcomes of Canadian children diagnosed with JIA between 2005 and 2010, an era when biologic medications were available.

Methods

The Research in Arthritis in Canadian Children Emphasizing Outcomes (ReACCh-Out) study was a multicentre inception cohort that prospectively recorded clinical outcomes of patients newly diagnosed with JIA in 2005–10. Follow-up was for a median of 3 years, concluding in May 2012 [6]. The clinical charts of children enrolled in the ReACCh-Out cohort at two centres (British Columbia Children’s Hospital in Vancouver, BC; and McMaster Children’s Hospital in Hamilton, ON) were reviewed in 2018 to ascertain disease status at the last available visit; prior to discharge or transition to adult care. Charts were identified through the ReACCh-Out records and de-identified data were collected in a case report form that was pilot tested in a sample of charts. The study was approved by research ethics boards at both institutions (UBC: H18-00628; McMaster: 4801).

We collected chart data on swollen and tender joints, current treatments, articular and extra-articular complications, imaging findings and ESR/CRP values. Disease status was determined according to Wallace criteria [7] as active or inactive. Clinical records of previous visits were evaluated to determine whether the patient met criteria for remission off medications (≥12 continuous months with inactive disease while being off all arthritis and uveitis medications) or remission on medications (≥6 continuous months with inactive disease while receiving medications). Joint damage was defined as the presence of a joint erosion or joint space narrowing in at least one joint, reported on radiographs or MRI studies done for clinical indication. Descriptive statistical analysis was performed with STATA version 12 (STATA Corp, College Station, TX, USA).

Results

A total of 247 out of 254 eligible patients (97%) were included. Seven children were excluded: two changed residence at an early age, one was lost to follow-up and in four the diagnosis changed (one each with pigmented villonodular synovitis, vascular malformation of the knee, Aicardi-Goutières syndrome and non-inflammatory musculoskeletal pain). The baseline characteristics of included patients at the time of recruitment into ReACCh-Out were typical of Western inception cohorts, except that enthesitis-related arthritis (ERA) was more frequent than RF-negative polyarthritis (Supplementary Table S1, available at Rheumatology online); 142 (58%) were female and the median (interquartile range (IQR)) age at last visit was 16.9 years (12.9–18.0). Median (IQR) follow-up was 5.6 years (3.3–8.3) after diagnosis with a maximum of 13.1 years. At the time of the last follow-up visit, 55% of patients had reached the age for transfer to adult care, 21% had been discharged from clinic and 23% were still being followed.

According to Wallace criteria, 180 patients (73%, CI 67, 78) had inactive disease. Of these, 116 (47%, CI 41–53) were in remission off medications and 61 (25%, CI 20, 30) were in remission on medications (Table 1). Three patients had inactive disease but did not fulfil criteria for remission. Patients with systemic JIA had the highest frequency of remission off medications (70%) and patients with RF-positive polyarthritis had the lowest (18%) (P < 0.05 by Fisher’s exact test for differences across categories). Among 67 patients with active disease at the last visit, 15 (22%) had no active joints and 29 (43%) had only one active joint. Raised inflammatory markers were observed in 12 patients (18%).

Table 1.

Selected outcomes of children with juvenile arthritis at last available visit

Group	Age in years	Years since diagnosis	Active disease	Remission on meds	Remission off meds	Still on treatment	Joint damage *
Oligoarthritis (n=99)	14 (11, 17)**	8 (3, 10)	21 (21)	29 (29)	47 (47)	51 (51)	10 (10)
ERA (n=52)	18 (17, 18)	4 (3, 6)	18 (35)	7 (13)	27 (52)	25 (48)	10 (19)
RF-Neg polyarthritis (n=39)	18 (13, 18)	5 (4, 8)	16 (41)	10 (26)	12 (31)	24 (62)	6 (15)
Systemic (n=23)	17 (13, 18)	6 (4, 8)	2 (9)	5 (22)	16 (70)	7 (30)	4 (17)
Psoriatic (n=10)	18 (17, 18)	5 (4, 6)	2 (20)	3 (30)	5 (50)	5(50)	6 (60)
RF-pos polyarthritis (n=11)	18 (15, 19)	6 (4, 10)	4 (36)	5 (45)	2 (18)	8 (73)	5 (45)
Undifferentiated (n=13)	17 (10, 18)	6 (5, 8)	4 (31)	2 (15)	7 (54)	6 (46)	4 (31)
All (n=247)	17 (13, 18)	6 (3, 8)	67 (27)	61 (25)	116 (47)	126 (51)	45 (18)

^*An erosion or joint space narrowing identified in at least one joint on x-rays or MRI done for clinical indication. There was no expectation or protocol for routine imaging of patients.

^**Numbers are median (IQR) rounded to the closest integer, or number (%) as appropriate.

At the last available visit, 51% of patients were on at least one anti-rheumatic medication: 10% were on NSAIDs ± joint injections, 19% were on DMARD ± NSAIDs ± joint injections and 22% were on biologics ± other treatments (Fig. 1).

Fig. 1.

Disease status at the last available visit

The clinical charts of 247 children with juvenile idiopathic arthritis diagnosed 2005–10 at two Canadian centres were reviewed in 2018, at the last available visit before discharge or transition to adult care. Active disease, remission on medications and remission off medications were defined as per Wallace criteria [7]. JI: joint injections with corticosteroids.

Uveitis was diagnosed in 24 patients (10%), of whom 10 (4%) had active uveitis requiring corticosteroid eye drops at the last visit. Two patients with oligoarthritis had uveitis-related complications, one developed band keratopathy and decreased visual acuity and the other developed a cataract. Micrognathia or facial asymmetry was observed in 11 patients (4%) and leg-length discrepancy in 12 patients (5%). Joint damage (at least one erosion or joint space narrowing in X-rays or MRI) was reported in 45 patients (18%), two patients had avascular necrosis and five (2%) required joint surgery, including two joint replacements.

Changes in ILAR category were observed in 31 patients (13%). Eight patients (3%) initially categorized as unclassified were assigned a definitive JIA category by the last clinic visit. Among 99 patients with oligoarthritis at enrolment, 14 (14%) had an oligoarthritis extended course, six (6%) were eventually diagnosed with IBD-associated arthritis and four (4%) with PsA. In retrospect, three of the six patients subsequently diagnosed with IBD-associated arthritis had some gastrointestinal symptoms at the time of diagnosis of JIA. None of these six patients received early biologics, but one was eventually treated with adalimumab.

Analyses repeated on the 138 patients who were >16 years at the last visit showed essentially the same results, except that 30% instead of 27% had active disease, but oligoarthritis was under-represented and ERA over-represented in this subsample.

Discussion

Relative to historical cohorts, young adults who were diagnosed with JIA early in the biologic era at two Canadian centres in this study had higher rates of remission and lower rates of permanent joint damage (<1% required joint replacements). Patients with systemic JIA had higher rates of remission, and patients with oligoarthritis had lower rates of progression to extended oligoarthritis. However, one in four young adults with JIA still had active disease and half were still receiving anti-rheumatic medications.

Perhaps the most comparable Canadian study from the pre-biologic era is the study by Oen et al. [3]. Among patients diagnosed 1974–94, 36% of patients with systemic, 52% with pauciarticular and none with RF-positive polyarticular onset JRA were in remission by the age of 16 years [3]. A prospective multicentre Nordic study of patients diagnosed between 1997 and 2000 reported that 42% of JIA patients were in remission off medications 8 years after disease onset [8]. Similar to our study, children with polyarthritis had the worst disease course. This is consistent with the 3-year outcomes reported by Selvaag et al. [9] who found that no patients with RF-positive polyarthritis and 16% with RF-negative polyarthritis were in remission off medications.

Our study found patients with systemic JIA had a good long-term prognosis with 70% in remission off medications and only 9% with active disease at a median of 5.7 years after diagnosis. Nordal et al. [8] similarly reported that 83% of patients with systemic JIA were in remission 8 years after disease onset. A more recent study of patients treated with anakinra as first-line therapy reported that 75% had inactive disease off medication 5 years after diagnosis [10]. Before the biologic era, 40% of children with systemic JIA had a monophasic course [11], but only 36% were in complete remission by the age of 16 years [3].

Oligoarthritis is usually considered a JIA category with good prognosis. However, 51% of children in our study remained on treatment at the last follow-up visit. Selvaag et al. [12] similarly reported that 50% had active disease at 5-year follow-up and that 30% had persistent disease at 30-year follow-up. Interestingly, we observed a lower rate of progression to extended oligoarthritis (14%) than in patients diagnosed at British Columbia Children’s Hospital 1965–94 (39.5%) [13]. Early treatment with biologics is unlikely to explain the low frequency of oligoarthritis extension. In patients with oligoarthritis in the ReACCh-Out cohort, the chance of receiving biologics in the first year after diagnosis was only 1.5% [6]. Biologics were eventually used in 13/99 children with oligoarthritis (13%), mostly to treat refractory uveitis or because they had an extended course. Early use of MTX is a more likely explanation for the low frequency of extension.

In our study, 10% of patients developed uveitis and 1% had major uveitis complications. A recent prospective German study also reported a 1% incidence of major ocular complications (nine cataracts and one band keratopathy among 954 JIA patients) [14].

The main strengths of our study are that patients were prospectively recruited and outcomes were known for 97% of patients. The main limitation is that data collection at the last available visit was retrospective. For example, a formal physician global assessment of disease activity using a visual analogue scale was often not available and statements that the disease was inactive, there were no active joints and the fact the child was on no treatment were considered to categorize the child as in remission off medications.

Conclusion

This long-term follow-up of an inception cohort diagnosed early in the biologic era suggests a reduction in permanent damage across JIA categories, a more favourable prognosis for systemic JIA and a lower progression to oligoarthritis extended category than in historical cohorts. However, one in four patients still had disease activity in early adulthood and one in two were receiving anti-rheumatic medications.

Supplementary data

Supplementary data are available at Rheumatology online.