When and in whom to stop etanercept after successful treatment of Juvenile Idiopathic Arthritis

Juvenile Idiopathic Arthritis (JIA) is the most common cause of chronic
arthritis in childhood. The term JIA encompasses all forms of arthritis that
begin before the age of 16 years, persist for more than 6 weeks, and are of
unknown cause. It is a heterogeneous disease comprising several disease
subtypes. Evidence is accumulating that early disease control is important to
prevent joint destruction, growth deformities and even blindness (from
JIA-associated chronic uveitis). Therefore, JIA therapy has changed in favour
of the early introduction of Disease Modifying Anti-Rheumatic Drugs (DMARDs)
and biologicals.
Since its introduction in 1999 etanercept, a TNF-alpha-blocker, has become an
important treatment for patients with refractory JIA. It is currently the most
frequent prescribed biological, and is proven to be effective in JIA patients
who previously had not respond to DMARDs, including methotrexate. (Lovell 2000
en Prince 2009) However, despite this success there are also concerns regarding
the long-term use of etanercept, since it suppresses the immune system of young
children.
After reaching remission, it is a logical step to try to reduce or discontinue
etanercept in order to prevent unnecessary side-effects and costs and to reduce
the burden of weekly injections. Unfortunately, no guidelines on when or how to
stop etanercept therapy are currently available and little is known about the
course of the disease in JIA patients after discontinuation.
Only one study (our pilot study) has been published on the subject when and how
to stop etanercept after successful treatment of JIA patients (Prince 2009).
All 19 JIA patients from the ABC-register who discontinued etanercept because
of a sustained good clinical response were evaluated. Ten of these patients
retained remission during follow up during a median of 0.8 years. All patients
who retained remission fulfilled the Wallace remission criteria at time of
discontinuation in contrast to only one third of the patients who flared.
Patients with a longer remission period and carefully tapering had a better
chance retaining remission. Nine patients flared after discontinuation of
etanercept, five within the first 6 months and 3 more in the following 3 months
after discontinuation. All eight patients who restarted etanercept after
disease flare regained a good clinical response which is reassuring. However
due to the fact that these data are from an observational study we can not
exclude that reasons related to patient and disease characteristics have
influenced the decision to stop etanercept at a certain moment causing
substantial bias. There were not enough patients in the study to evaluate other
predictors for successful stopping. This underlines the need for a study with
an interventional design and more patients to answer the research questions.
The National Institute for Clinical Excellence (NICE), UK, recommend in their
guidelines a 2 years disease-free period before discontinuation of etanercept
(NICE 2005). However, it is not clear on what data they based their
recommendation.
Foell et al. published data on when to stop methotrexate (MTX) after successful
treatment and at the same time they investigated the usefulness of myeloid
related proteins 8 and 14 (MRP8/ MRP14) (Foell 2004). Conclusions were that
residual synovial inflammation seemed to influence the rate of disease flares
and that MRP in clinical inactive arthritis may help to identify patients in
whom MTX can be safely withdrawn. But they also mention that they cannot
exclude the possibility that reasons related to patient characteristics
influenced the decision for MTX earlier or later, which might have influenced
the subsequent occurrence of flares.
Also, data on Rheumatoid Arthritis (RA) treatment on this subject are limited.
Miyamura et al. published results on two patients who had prolonged clinical
and radiographic remission of RA after the discontinuation of etanercept
(Miyamura 2010). Brocq et al. reported on 21 RA patients in remission who were
taken of anti-TNF-alpha (14 of etanercept) (Brocq 2009). Only 25% of patients
retained remission. Patients still in remission had a longer mean period of
anti-TNF-alpha use and longer mean period of remission before discontinuation.
Concluding from published data; no consensus has been established concerning
the important clinical question when and in whom to stop etanercept therapy in
JIA after successful treatment.

Hypothesis:
A substantial proportion of the JIA patients in remission (according to the
Wallace criteria) will be able to discontinue etanercept successfully.

Goals:
To investigate in a randomized controlled trial:
- which proportion of JIA patients in remission can successfully discontinue
etanercept compared to JIA patients in remission who continue etanercept;
- if time in remission on etanercept is an important factor in retaining
remission after discontinuation of etanercept;
- if flare-rate is different between the subgroups 3-9 months and 9-18 months
in remission on etanercept.

To evaluate
- predicting factors (patient or disease characteristics (including time in
remission) and MRP8/MRP14) for successfully discontinuing etanercept;
- the disease course after discontinuation of etanercept therapy (time to
flare) and the effect of restarting etanercept after flaring.

The study consists out of 2 parts:
1. A randomized controlled trial. Intervention: discontinuation etanercept
2. An observational study of all patients discontinuating etanercept.

1. A randomized controlled trial (intervention: discontinuation of etanercept)
All eligible patients selected from the ABC-register and during a 12 month
inclusion period will be stratified into 2 subgroups (3-9 and 9-18 months in
remission under etanercept treatment) and than randomized in 2 arms; an
intervention and a control group. Stratification in these 2 subgroups will
ensure an equal representation of time in remission in both arms and therefore
avoid confounding. We will test for interaction between the 2 subgroups. Both
subgroups (3-9 and 9-18 months in remission under etanercept treatment) will be
analyzed together, with adjustment for time in remission.
Patients from the ABC-register who already meet the in- and exclusion criteria
before beginning of the study will be included directly after approval of the
Medical Ethical Committee and depending on their time in remission enrol in the
different subgroups.
Patients meeting the in- and exclusion criteria during the inclusion period,
will be included as soon as these criteria are met.
The study will be open (not blinded for both patients and physicians).
Randomization, in block sizes of 6 patients, will be coordinated from the
Erasmus MC Sophia Rotterdam, and will be confirmed by email or fax within 24
hours.
Duration of this randomized part will be 9 months in total; 3 months tapering
off and 6 months follow-up from discontinuation. We consider a follow-up of 6
months from discontinuation long enough, since in our pilot study more than 50%
of the patients flared within 6 months.

Intervention: Patients will first taper the etanercept medication. If patients
are on a dosing schedule of twice a week 0.4 mg/kg (with a maximum of 50
mg/week) frequency will be lowered to once a week 0.4 mg/kg (with a maximum of
25 mg/week) for 3 months. If patients are on a dosing schedule of once a week
0.8 mg/kg (with a maximum of 50 mg/week) frequency will be lowered to once a
week 0.4 mg/kg (with a maximum of 25 mg/ week) for 3 months. If patients don*t
experience a flare during tapering of etanercept it will be discontinued
completely 3 months after start of the study. After discontinuation of
etanercept, patients will be followed for another 6 months for this study part.
Concomitant medication: MTX, if used, will be kept stable through the study
course, with adjustment for growth allowed. In case of complaints and not
meeting the criteria of a flare, it is allowed to start or raise the dosage of
NSAIDs. No corticosteroids (including intra-articular corticosteroids up to 6
months prior to inclusion) and other synthetic and biologic DMARDs are allowed.

Control: Patients will continue the same dosage of etanercept, as it is at
start of the study. Follow-up: 9 months.
The dosage of etanercept and MTX, if used, will be kept stable through the
study course, with adjustment for growth allowed. In case of complaints and not
meeting the criteria of a flare, it is allowed to start or raise the dosage of
NSAIDs. No corticosteroids (including intra-articular corticosteroids up to 6
months prior to inclusion) and other synthetic and biologic DMARDs are allowed.

2. An observational study of all patients discontinuating etanercept
(predictive factors for successful discontinuation)
Patients, who were randomized into the control group (continuation of
etanercept) and are still in remission at the end of part 1, will discontinue
etanercept in the same way as the intervention group in part 1 did. All
patients who discontinue etanercept will be followed for 15 months in total .
All patients who start tapering off etanercept will be included for this
analysis.
We will analyze the following predictive factors for successful discontinuation
of etanercept:
- Duration of remission till withdrawal of etanercept (continuous variable)
- MTX comedication at inclusion of the study (yes or no)
- Optimal dosage of MTX given before start of etanercept? (yes or no, optimal
dosage MTX defined as 15-20 mg/m2/week, with a maximum of 25mg/week)
- Subtype JIA (systemic vs non-systemic patients)
- Immunological parameters (f.e. MRP8/ MRP14, continuous variable)
- Presence of pre-existent radiological damage before start of etanercept
therapy (yes or no)

Flare:
If the local investigator notices (in both study parts) that a patient shows an
increase in disease activity, the flare-criteria will be checked and verified
by the research physician. Patients will be treated according to the judgement
of the treating physician and etanercept therapy can immediately be
reintroduced. Response to reintroduction of etanercept will be measured
according the ACR pediatric 30, 50 and 70 improvement criteria. If a patient or
parent re-introduces etanercept, although the criteria are not completely
fulfilled, it is considered as protocol violence.
We consider a patient in remission to flare if one or more of the following
occurs (=flare criteria):
- active arthritis in two or more joints;
- ESR > 30 mm/hour not otherwise explained;
- physician*s global of 30mm or more;
- active uveitis;
- fever, rash, serositis or generalized lymphadenopathy (for sJIA patients
only);
- signs of SI and spine involvement according the treating physician (for ERA
patients only);

Follow-up:
Fixed study visits are at start of the study, during tapering of etanercept,
and regularly after discontinuation of etanercept (see protocol).

Intervention: withdrawal of etanercept therapy.

Control: continuation of therapy (etanercept).

The risk of withdrawal of etanercept is the risk of exacerbation of the
disease. However with the continuation of treatment there also remains a
(small) risk of exacerbation. In addition there are more risks related to
continuation of etanercept, especially with the long-term use of etanercept
iwith an increased risk of infection, and possible increased risk of the
development of other autoimmune diseases and even malignancies. It is
reassuring that in our pilot study all patients in whom etanercept was
restarted due to an exacerbation, responded well to the reintroduction of
etanercept.

The burden of the patients in this study is negligible because standard usual
care is applied.