Clinical remission in juvenile idiopathic arthritis after termination of etanercept

Biologicals are very effective for inhibiting disease progression in active juvenile idiopathic arthritis (JIA). To date, there have been no recommendations on how and when to stop therapy with TNF inhibitors. Our objective was to analyze characteristics and the disease course of JIA patients who discontinued etanercept due to achievement of inactive disease. Data of 39 patients with JIA from two clinical pediatric rheumatology centers in Bydgoszcz and Lublin (Poland) were analyzed retrospectively. All patients discontinued etanercept due to a remission on treatment. Etanercept was started after a mean 33.7 ± 36 (range 3–137) months of disease. The mean duration of therapy with etanercept was 34.7 ± 16.7 (range 6–72) months, with a mean duration of remission on medication 21.3 ± 9.6 (range 4–42) months before withdrawal of etanercept. The mean duration of remission after etanercept discontinuation was 14.2 ± 12.1 (range of 1–60) months. Only 12/39 (30.8 %) patients did not develop a disease exacerbation until the end of the study. Early flares, that is less than 6 months after termination of etanercept, were observed in 15/39 (38.5 %) patients. Twelve (30.8 %) patients restarted etanercept after exacerbation—all patients responded satisfactorily. Our data show that etanercept discontinuation in a substantial proportion of JIA patients results in early disease exacerbation. In many cases, reintroduction of etanercept is needed. Patients, in whom etanercept was restarted, responded satisfactorily.


Introduction
Juvenile idiopathic arthritis (JIA) is the most common rheumatic disease of childhood. The essence of the disease is a chronic autoimmune inflammation. The main therapeutic goal should be to achieve the reduction of inflammation and initial improvement, then arriving at remission of disease.
For many patients to achieve inactive disease is difficult. In spite of treatment applied, in some patients, disease activity is continued into adulthood [1,2].
Following a better understanding of the pathogenesis of JIA, in the last 20 years, the principles of pharmacotherapy in JIA have changed. Initially, synthetic and now also biological disease-modifying drugs (DMARDs) are the main element of the therapy. Early initiated, intensely combined treatment creates the possibility of achieving inactive disease (ID) status and its remission.
The preliminary criteria for clinical remission, for selected categories of juvenile idiopathic arthritis, were described by Wallace et al. [3] in 2004.
Experience of over a decade of TNF inhibitors use in JIA allowed formulating the criteria for inclusion treatment in JIA. So far, there have been recommendations on how and when we should stop TNF inhibitor therapy. The questions are: how long should one treat the problem biologically? What factors influence the duration of remission after biological treatment? Should biological drug be discontinued suddenly or gradually by reduction of the dose or should one prolong the intervals between doses?
Six-month disease inactivity is required for fulfilling the criteria for inactive disease on medication, and 12 months for the criteria for inactive disease off medications.

Objective
We analyzed the disease course of JIA patients that discontinued etanercept due to achievement of the stage of inactive disease.

Materials and methods
The data of 39 patients (12-male, 27-female), with JIA from two clinical pediatric rheumatology centers in Lublin and Bydgoszcz (Poland), were analyzed retrospectively. All patients discontinued etanercept due to a remission on treatment. Inactive disease was defined according to the preliminary criteria of Wallace et al. [3]. The clinical subtypes of JIA were systemic-7, polyarthritis-14, oligoarthritis-14, enthesitis-related arthritis-3, psoriatic arthritis-1.
Patients were treated with etanercept 2 times a week, 0.4 mg per kg of body weight per dose, until the end of therapy. Treatment was terminated abruptly.

Results
Etanercept was started after a mean 33.7 ± 36 (range 3-137) months of disease. The mean duration of therapy with etanercept was 34.7 ± 16.7 (range 6-72) months, with a mean duration of remission on medication 21.3 ± 9.6 (range 4-42) months before withdrawal of etanercept.
The mean duration of remission after etanercept discontinuation was 14.2 ± 12.1 (range 1-60) months. 12/39 (30.8 %) patients, until the end of the study, did not develop a disease exacerbation and remained in long-term remission off medication-a mean 25.4 ± 12 (range 16-60) months. Because of exacerbation of the disease, 24 children required treatment with methotrexate, 5 with cyclosporine A, 1 patient required treatment with hydroxychlorochine. Early flares, that is less than 6 months after termination of etanercept, were observed in 15/39 (38.5 %) patients.
Twelve (30.8 %) patients restarted etanercept after exacerbation, due to lack of improvement after no biological DMARDs. All patients in whom etanercept was reinitiated responded satisfactorily.
Statistical analysis did not show any correlation between gender, age and type of onset, duration of disease before etanercept introduction to therapy, duration of treatment with etanercept or sustaining remission off medication.

Discussion
Only in some scarce reports, the duration of etanercept therapy was being investigated, and the question how to terminate etanercept therapy for inactive disease in JIA was tackled. The groups of patients presented till now and quoted below were less numerous than as we describe it.
The mean duration of etanercept therapy in our group of children was 34.7 months. Prince [6], the mean duration on therapy with etanercept was shorter-that is about 19 months.
In a cohort of 483 etanercept-treated JIA patients monitored in the British Register [7], a total of 100 (20.7 %) patients discontinued etanercept; only 9 due to disease control. The authors report that in their group followed for a median of 2 years (maximum 5 years), the majority (69 %) remain on the drug.
In our patients, the mean time of remission after discontinuation of etanercept was 14.2 months, although it should be noted that in as many as 15/39 children (38.5 %), exacerbation occurred in less than 6 months.
Prince et al. [4] analyzed the clinical course of the disease in 19 patients with JIA after discontinuation of etanercept because of sustained good clinical response. In ten of them, remission lasted over a median of 0.8 years. It was found that the longer clinical remission off etanercept therapy was associated with previous longer clinical remission on etanercept. In the remaining 9 patients, exacerbation occurred earlier than 0.8 years. It was noted that in four of them, etanercept was withdrawn abruptly. The authors suggest the beneficial effect of prolonged duration of therapy during clinical remission on medication. They also suggest tapering the etanercept dose carefully. Prince et al. [4] believes that patients with JIA should meet the criteria of clinical remission on etanercept medication for at least of 1.5 years before considering discontinuation of etanercept.
Our study did not confirm any correlation with demographic characteristics, clinical symptoms and a course of treatment.
Pratsidou-Gertsi et al. [5] evaluated the course of the disease in 11 patients with JIA (9-polyarthritis, 2-oligoarthritis disease course) who discontinued etanercept after achieving remission. The Juvenile Arthritis Disease Activity Score (JADAS) was used to grade the JIA activity Remesal et al. [6] presented interesting results of a retrospective analysis of 26 patients with JIA who discontinued therapy with etanercept due to the inactivation of the disease. The clinical subtypes of JIA were: 11 cases of enthesitis-related arthritis, 7-rheumatoid factor-negative polyarthritis cases, 2 cases of systemic JIA with polyarticular involvement, 1 psoriatic arthritis and 1 persistent oligoarthritis. Fourteen patients withdrew from therapy with etanercept abruptly, in 12 either by reducing the dose or by increasing the interval between doses. Etanercept was restarted on all relapsed patients. The mean treatment duration was 19 ± 8.4 (9.6-38.5) months. Mean duration of inactive disease before etanercept discontinuation was 14.7 ± 8.6 (1-36) months. Eighteen cases (69 %) relapsed at a mean of 5.8 ± 5.3 (0.6-15.9) months after drug discontinuation, whereas in the other 8 (31 %) patients, the disease remained inactive for a mean of 21 ± 14.7 (range 5-44.5) months. In 9 patients, after withdrawal of etanercept, the disease remained inactive for at least 12 months. In 4 of them, it exacerbated from 1.5 to 4 months later, and the remaining 5 continued in remission for a mean of 17 ± 13 (range 1.1-32.5) months until the end of the study. It was shown that cumulative probability of maintaining the disease in the inactive state for 6 months after treatment with etanercept is characteristic of 50 % of patients and for 12 months-of 39 %. There was no significant difference in duration of remission between the groups of patients in whom etanercept was discontinued abruptly or the dose was reduced gradually (11 vs. 14 months p = 0.48). There was no correlation between the duration of remission on etanercept and the duration of the remission after etanercept discontinuation. All patients in whom etanercept was restarted responded satisfactorily. The authors conclude that majority of patients (69 %) relapsed after etanercept discontinuation. The probability of maintaining the disease in asymptomatic stage for 6 months after termination of therapy is 50 %. In contrast to the results of Prince et al. [4], this analysis did not achieve significant correlation between duration of clinical remission on etanercept therapy and duration of the remission after discontinuation.
It is interesting that despite the shorter duration of etanercept treatment and shorter duration of clinical remission on therapy, in Remesal et al. [6] observation, we found a similar proportion of relapses after etanercept treatment (about 69 %).
Incidence of early exacerbations, that is up to 6 months, is varied. In our study group, the exacerbations up to 6 months occurred in 15/39 (38.5 %). In group of Remesal et al. [6] in 18/26 (69 %), remission was maintained in the mean time of 5.8 ± 5.3 months. In the small group presented by Prince et al. [4], this proportion was 4/19 (21.1 %), while in the similar group of Pratsidou-Gertsi et al. [5], the mean duration of the remission off etanercept was only 3 months.
In 12/26 patients of Remesal et al. [6], the dose of etanercept was reduced gradually. There was no exacerbation of the disease until complete withdrawal, which might suggest that low doses of etanercept may be effective to maintain remission. Prince et al. [5] also suggest gradual reduction of the dose of etanercept. The authors pointed out that early exacerbations, that is less than 0.8 years, were observed in 4/19 (21 %) children whose therapy was finished abruptly. Our observations and the observations of other authors show a high proportion of early exacerbations. However, all patients who had restarted etanercept responded satisfactorily.
It should be emphasized that biological treatment is used in patients with the most active course of the disease, not responding to methotrexate or other DMARDs. Probably, the majority (2/3) of these patients require a permanent or much longer etanercept therapy than quoted in this paper.

Conclusions
Our study does not show correlation between the duration of remission off etanercept with duration of total etanercept treatment duration, clinical remission on etanercept, type of disease onset and duration of disease to introduction etanercept.
Discontinuation of etanercept therapy in a substantial proportion of JIA patients resulted in early exacerbation of the disease.
In many cases, reintroduction of etanercept is needed. Patients, in whom etanercept was restarted, responded satisfactorily.
Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.