Factors Related to Sustained Discontinuation of Medications for Well-Controlled JIA in the Childhood Arthritis & Rheumatology Research Alliance Registry

Daniel B. Horton1,2, Fenglong Xie3, Melissa L. Mannion3, Sarah Ringold4, Colleen K. Correll5, Anne Dennos6, Timothy Beukelman3 for the CARRA Registry Investigators

1Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ, USA

2Rutgers Institute for Health, Health Care Policy and Aging Research, New Brunswick, NJ, USA

3University of Alabama at Birmingham, Birmingham, AL, USA

4Seattle Children's Hospital, Seattle, WA, USA

5University of Minnesota, Minneapolis, MN, USA

6Duke Clinical Research Institute, Durham, NC, USA

Background: Stopping medications is a priority for many patients with well-controlled JIA, but few factors predict favorable outcomes after discontinuation. We examined factors associated with sustained discontinuation off disease-modifying drugs in a large pediatric registry.

Methods: We conducted a case-control study using the Childhood Arthritis & Rheumatology Research Alliance Registry ofclinical data from >55 pediatric rheumatology clinics in the United Statesand Canada. The study included children with JIA who started at least 1 conventional or biologic DMARD and had at least 2 years of available subsequent data. Reasons for drug discontinuation were obtained from the medication log. Cases with sustained discontinuationhad at least 1 year of drug-free follow-up after discontinuation for well-controlled disease. Comparators included children who stopped DMARDs for at least 30 days and restarted within 1 year (control group 1) and those who did not discontinue all DMARDs for well-controlled disease for ≥30 days(control group 2). Children in each group may have stopped DMARDs for reasons besides well-controlled JIA.We excluded children with <1 year of follow-up off medicines without restarting DMARDs. Characteristics ofcases and controls were compared using descriptive statistics.P-values were calculated by Wilcoxon rank sum testing for continuous variables and by chi-square or Fisher's exact test for categorical variables.

Results:There were 1338 children with JIA in the study who started DMARDs, of whom 545 (41%) stopped all drugs due to well-controlled diseaseafter median 1.9 years of DMARD use (interquartile range [IQR] 1.1, 3.2). Of these, 222(41%) had sustained discontinuation for ≥1 year (cases) and 323 (59%) restarted DMARDs within 1 year (control group 1) (Table).In control group 1 with unsustained discontinuation, only 24 children (10% of control group 1, or 2% of the total cohort) subsequently stopped all DMARDs for ≥1 year.Among all childrenin the study who started a DMARD, cases with sustained discontinuationwere younger at diagnosis (P < 0.01), more likely to have systemic JIA and less likely to have RF+ polyarticular JIA (P = 0.03), less likely to have complications from uveitis (P < 0.01), and less likely to have used a biologic DMARD (<0.01). Compared with controls with unsustained discontinuation, cases were less likely to have complications from uveitis (P = 0.05) or have used biologics (<0.01).Time to diagnosis, history of uveitis, radiographic joint damage, and conventional DMARD use were not associated with sustained discontinuation in either comparison.

Conclusion: In a large multicenter cohort of children with JIA on DMARDs, only 1 in 5 children stopped DMARDs for well-controlled disease for at least 1 year. Lack of uveitis complications and exclusive use of conventional DMARDs were associated with sustained discontinuation.

Table. Characteristics of study subjects with JIA who started at least 1 DMARD
Characteristic (N, % unless otherwise noted) / Cases: off DMARDs≥12months1 (N=222) / Control group 1: off DMARDs <12 months1 (N=323) / Control group 2: not off DMARDs >30 days (N=793) / P-value2 (cases vs. control groups 1 and 2) / P-value2(cases vs. control group 1)
Demographic
Age at diagnosis (years), median (IQR) / 4.0 (2.0, 8.0) / 4.0 (2.0, 8.0) / 6.0 (2.0, 11.0) / <0.01 / 0.70
Time from symptom onset to diagnosis (days), median (IQR) / 47 (29, 86) / 58 (26, 92) / 54 (29, 132) / 0.49 / 0.61
Time from diagnosis to start of first DMARD (days), median (IQR) / 37 (1, 201) / 28 (1, 230) / 34 (1, 350) / 0.59 / 0.93
Time from start of DMARD to first DMARD discontinuation (years), median (IQR) / 1.9 (1.2, 3.3) / 1.9 (1.1, 3.1) / - / - / 0.51
Time from start of first DMARD to end of follow-up (years), median (IQR) / 6.0 (4.4, 8.8) / 5.9 (3.9, 8.6) / 4.2 (2.8, 6.7) / <0.01 / 0.26
Female sex / 160 (72%) / 253 (78%) / 605 (76%) / 0.13 / 0.09
White race / 187 (84%) / 272 (84%) / 646 (82%) / 0.47 / 0.99
Latino ethnicity / 22 (13%) / 22 (8%) / 83 (12%) / 0.83 / 0.46
Residence in US / 217 (98%) / 314 (97%) / 770 (97%) / 0.61 / 0.70
Clinical
JIA category / 0.03 / 0.14
Systemic / 33 (15%) / 30 (12%) / 107 (14%)
RF+ polyarticular / 11 (5%) / 19 (6%) / 98 (12%)
Other / 178 (80%) / 272 (85%) / 588 (74%)
≥5 total joints affected / 193 (87%) / 295 (91%) / 719 (91%) / 0.15 / 0.23
Radiographic evidence of joint damage / 31 (14%) / 56 (17%) / 117 (15%) / 0.55 / 0.28
ANA positive / 96 (43%) / 139 (43%) / 334 (42%) / 0.81 / 0.96
History of uveitis / 19 (9%) / 19 (6%) / 80 (10%) / 0.86 / 0.23
Uveitis complications / 0 / 7 (2%) / 22 (3%) / <0.01 / 0.05
Any conventional DMARD use / 201 (91%) / 287 (89%) / 719 (91%) / 0.86 / 0.53
Any biologic DMARD use / 84 (38%) / 174 (54%) / 691 (87%) / 0.01 / <0.01
ANA, antinuclear antibody; DMARD, disease-modifying antirheumatic drug; IQR, interquartile range; RF, rheumatoid factor.
1 Refers to discontinuation from all DMARDs for at least 30 days for well-controlled JIA
2P-values calculated by Wilcoxon rank sum testing for continuous variables and chi-square or Fisher's exact test for categorical variables

Ethics Approval

The study was approved by University of Alabama's Institutional Review Board, protocol number X170112004.

Funding

Funding for this work came from a CARRA-AF Publication Support Grant.