Articles

Intra-articular corticosteroids versus intra-articular

corticosteroids plus methotrexate in oligoarticular juvenile
idiopathic arthritis: a multicentre, prospective, randomised,
open-label trial
Angelo Ravelli, Sergio Davì, Giulia Bracciolini, Angela Pistorio, Alessandro Consolaro, Evert Hendrik Pieter van Dijkhuizen, Bianca Lattanzi,
Giovanni Filocamo, Sara Verazza, Valeria Gerloni, Maurizio Gattinara, Irene Pontikaki, Antonella Insalaco, Fabrizio De Benedetti, Adele Civino,
Giuseppe Presta, Luciana Breda, Valentina Marzetti, Serena Pastore, Silvia Magni-Manzoni, Maria Cristina Maggio, Franco Garofalo,
Donato Rigante, Marco Gattorno, Clara Malattia, Paolo Picco, Stefania Viola, Stefano Lanni, Nicolino Ruperto, Alberto Martini, for the Italian
Pediatric Rheumatology Study Group

Summary
Background Little evidence-based information is available to guide the treatment of oligoarticular juvenile idiopathic Lancet 2017; 389: 909–16
arthritis. We aimed to investigate whether oral methotrexate increases the efficacy of intra-articular corticosteroid therapy. Published Online
February 2, 2017
http://dx.doi.org/10.1016/
Methods We did this prospective, open-label, randomised trial at ten hospitals in Italy. Using a concealed computer- S0140-6736(17)30065-X
generated list, children younger than 18 years with oligoarticular-onset disease were randomly assigned (1:1) to intra-
See Comment page 883
articular corticosteroids alone or in combination with oral methotrexate (15 mg/m²; maximum 20 mg). Corticosteroids
Istituto Giannina Gaslini,
used were triamcinolone hexacetonide (shoulder, elbow, wrist, knee, and tibiotalar joints) or methylprednisolone Genoa, Italy (Prof A Ravelli MD,
acetate (ie, subtalar and tarsal joints). We did not mask patients or investigators to treatment assignments. Our S Davì MD, A Pistorio MD,
primary outcome was the proportion of patients in the intention-to-treat population who had remission of arthritis A Consolaro MD,
E H P van Dijkhuizen MD,
in all injected joints at 12 months. This trial is registered with European Union Clinical Trials Register, EudraCT
M Gattorno MD, C Malattia MD,
number 2008-006741-70. P Picco MD, S Viola MD,
S Lanni MD, N Ruperto MD,
Findings Between July 7, 2009, and March 31, 2013, we screened 226 participants and randomly assigned 102 to Prof A Martini MD); Università
degli Studi di Genova, Genoa,
intra-articular corticosteroids alone and 105 to intra-articular corticosteroids plus methotrexate. 33 (32%) patients
Italy (Prof A Ravelli,
assigned to intra-articular corticosteroids alone and 39 (37%) assigned to intra-articular corticosteroids and G Bracciolini MD, A Consolaro,
methotrexate therapy had remission of arthritis in all injected joints (p=0·48). Adverse events were recorded for S Verazza MD, C Malattia,
20 (17%) patients who received methotrexate, which led to permanent treatment discontinuation in two patients Prof A Martini); Azienda
Ospedaliera Universitaria
(one due to increased liver transaminases and one due to gastrointestinal discomfort). No patient had a serious
Ospedali Riuniti Salesi
adverse event. Children’s Hospital, Ancona,
Italy (B Lattanzi MD);
Interpretation Concomitant administration of methotrexate did not augment the effectiveness of intra-articular Fondazione IRCCS Ca’ Granda
Ospedale Maggiore Policlinico,
corticosteroid therapy. Future studies are needed to define the optimal therapeutic strategies for oligoarticular juvenile Milan, Italy (G Filocamo MD);
idiopathic arthritis. Istituto Gaetano Pini, Milan,
Italy (V Gerloni MD,
Funding Italian Agency of Drug Evaluation. M Gattinara MD, I Pontikaki MD);
Ospedale Pediatrico Bambino
Gesù, Rome, Italy
Introduction (50–80%) of all white children with chronic arthritis in (A Insalaco MD,
Juvenile idiopathic arthritis comprises a heterogeneous North America and Europe.5 Although structural joint F De Benedetti MD,
group of disorders of unknown cause. Disorders are changes and functional disability in oligoarthritis are S Magni-Manzoni MD); Azienda
Ospedaliera Cardinale G Panico,
currently classified into seven categories on the basis of often less frequent and severe than those seen in other Tricase, Italy
the clinical and laboratory features present in the first forms of juvenile idiopathic arthritis, children with this (A Civino MD, G Presta MD);
6 months of illness.1 With a prevalence of approximately disease might develop significant musculoskeletal Ospedale Policlinico, Chieti,
one patient per 1000 population, it is one of the more abnormalities, such as fixed flexion contractures, valgus Italy (L Breda MD,
V Marzetti MD); IRCCS Burlo
common chronic diseases of childhood and a major deformities, and localised growth abnormalities of bones Garofolo, Trieste, Italy
cause of acquired disability.2 Many children with juvenile (which can, in turn, produce complications, such as leg- (S Pastore MD); Università degli
idiopathic arthritis do not achieve long-term remission, length inequality or micrognathia).2 Furthermore, Studi di Palermo, Palermo, Italy
thus the burden of disease to the patient, family, and 25–50% of patients have a progressive increase in the (M C Maggio MD); Ospedale
degli Infermi, Biella, Italy
ultimate society, is substantial.3,4 number of affected joints over time (so-called extended (F Garofalo MD); and Università
Oligoarticular juvenile idiopathic arthritis is defined as oligoarthritis), so that by 1–2 years after onset, they have Cattolica Sacro Cuore e
an arthritis that affects four or fewer joints during the polyarthritis (affecting five or more joints). In this Fondazione Policlinico
Universitario A Gemelli, Rome,
first 6 months of illness. It is a distinctly, if not uniquely, subgroup, structural joint damage is frequent and the
Italy (D Rigante MD)
paediatric disease, and accounts for the vast majority probability of remission is low.6,7

www.thelancet.com Vol 389 March 4, 2017 909

 Articles
Correspondence to:
Prof Angelo Ravelli, Pediatria II- Research in context
Reumatologia, Istituto Giannina
Gaslini, Genoa 16147, Italy Evidence before this study
angeloravelli@gaslini.org Oligoarthritis is the most common subtype of juvenile
idiopathic arthritis in white children. We searched PubMed
for papers published between Jan 1, 1980, to Dec 31, 2008,
using the keywords: “juvenile rheumatoid arthritis”, “juvenile
chronic arthritis”, “juvenile idiopathic arthritis”, “oligoarthritis”,
“oligoarticular”, “treatment”, “therapy”, and “trial”. Our search
results highlighted a paucity of randomised clinical trials
assessing the efficacy and safety of therapeutic interventions in
oligoarthritis.
Added value of this study
Concomitant administration of methotrexate might prolong
and, to a lesser extent, augment the effectiveness of
By contrast with the numerous randomised controlled
trials that have been done in polyarticular and systemic
juvenile idiopathic arthritis,8 little evidence-based
information is available to guide the treatment of
oligoarthritis.9–12 Additionally, most studies have included
non-steroidal anti-inflammatory drugs (NSAIDs), which
might alleviate inflammatory symptoms, but are rarely
able to control arthritis. As a result, the management of
this condition is largely empirical and is likely variable
among practitioners.13 The 2011 recommendations from
the American College of Rheumatology provide some
guidance for the initial treatment of oligoarthritis.14
However, it remains unclear when in the course of
disease and in what combinations these treatments
should be started to produce optimal outcomes.
Intra-articular corticosteroid injections are widely used
in the management of children with oligoarthritis to
induce short-term relief of inflammation symptoms and
functional recovery and to obviate the need for regular
systemic therapy.15,16 This intervention could enable
correction of joint contractures, prevention of leg-
length discrepancy, regression of Baker’s cysts, and
improvement of tenosynovitis.17 Many paediatric
rheumatologists use intra-articular corticosteroid therapy
as first-line therapy in oligoarthritis.15,16 However,
although intra-articular corticosteroid injections are
highly efficacious, most children have recurrence of
arthritis in injected joints (flare) after a variable period of
time.18 Arthritis flare might require a repeat injection of
corticosteroids, which can cause considerable emotional
distress to children and their parents. In addition, the
need of general anaesthesia in younger children can
increase organisational and financial costs.
Methotrexate is the cornerstone disease-modifying
antirheumatic drug for the treatment of juvenile
idiopathic arthritis on the basis of results from two
randomised controlled trials.19,20 However, these trials
have enrolled only patients with polyarthritis.19,20
Although this medication is widely used also in children
910 www.thelancet.com Vol 389 March 4, 2017
intra-articular corticosteroid therapy in children with
oligoarthritis, without an appreciable increase in toxicity.
The addition of methotrexate did not reduce the prevalence
of new onset of synovitis in previously unaffected joints.
Implication of all the available evidence
The combination of intra-articular corticosteroids and
methotrexate could be considered as reference treatment in
everyday clinical practice for paediatricians. Evaluating the
capacity of therapeutic interventions to prevent arthritis
extension in oligoarthritis should be the objective of future
clinical trials.
with oligoarthritis, typically after failure of NSAIDs or
intra-articular corticosteroids, its exact place in the
management of this subset is still unclear. In the
American College of Rheumatology recommendations,14
all advice about methotrexate initiation in children with
history of arthritis of four joints or fewer was assigned a
level of evidence C. The recent observation that early
methotrexate therapy might prevent the onset of uveitis
in juvenile idiopathic arthritis21 could support its first-line
use in children with oligoarthritis, who are at high risk
for this potentially serious complication.
There is therefore a need for evidence-based data to
help practitioners increase rationality in the therapeutic
approach to oligoarticular juvenile idiopathic arthritis.
Additionally, considering that current clinical practice
mandates the achievement of complete and sustained
disease control, it would be desirable to explore the role
of early aggressive protocols. In this respect, a key and
still unanswered question is whether the combination
with methotrexate increases and prolongs the
effectiveness of intra-articular corticosteroid therapy.
Additionally, non-controlled studies suggest that
concurrent methotrexate therapy reduces the risk of
arthritis flare after intra-articular corticosteroid
injection.22,23 We designed this trial aiming to assess
whether the concomitant administration of methotrexate
to children with oligoarticular juvenile idiopathic arthritis
who undergo intra-articular corticosteroid therapy
augments the rate and duration of arthritis remission,
without exposing children to an increased frequency of
untoward side-effects.
Methods
Study design and participants
We did a prospective, randomised, open-label trial at
ten hospital units that are part of the Italian Paediatric
Rheumatology Study Group. We enrolled children
younger than 18 years with oligoarticular juvenile
idiopathic arthritis (as per the International League of

Associations for Rheumatology [ILAR] criteria),1 who
were candidates to receive an intra-articular corticosteroid
injection in at least two or more joints. However, after an
amendment of the protocol (approved by both ethics
committee and Italian Medicines Agency) inclusion
criteria also comprised patients who received one
injection in a single target joint if it was a shoulder,
elbow, wrist, ankle, or a knee which had a flare of arthritis
within 12 months after a previous intra-articular
corticosteroid injection. Patients with knee monoarthritis
who received their first intra-articular corticosteroid
injection were excluded because this subgroup has the
highest rate of 12 month remission.18 Furthermore, most
participating investigators declared that they were not
willing to prescribe methotrexate to children with this
disease phenotype. Major exclusion criteria were for
previous treatment with methotrexate or other synthetic
or biologic disease-modifying antirheumatic drugs and
administration of systemic or intra-articular cortico­
steroids in the 3 months before enrolment. Continuing
treatment with NSAIDs did not lead to exclusion from
the study. However, this therapy had to be discontinued
at trial enrolment. The administration of live-attenuated
vaccines was not allowed for the duration of the trial.
Personnel at the coordinating centre at Gaslini Institute
in Genoa, Italy, verified patients’ inclusion and exclusion
criteria.
Local ethics committees at each participating centre
approved the study protocol. We obtained written
informed consent or assent from every parent or
guardian.
Randomisation and masking
Using SPSS version 19 to generate a progressive
sequential list with no restrictions, we randomly assigned
patients in a 1 to 1 ratio to either intra-articular
corticosteroids alone or intra-articular corticosteroids
plus methotrexate. To conceal assignments, the
randomisation list was accessible only by personnel at
the coordinating centre; participating investigators
received the final allocation via telephone or email.
Participants, clinicians (both treating clinicians and
outcome assessors), and statisticians were not masked to
group assignment.
Procedures
The preparation used for intra-articular corticosteroid
injections was triamcinolone hexacetonide for large
joints (shoulder, elbow, wrist, knee, and tibiotalar joints)
or methylprednisolone acetate for small hand and foot
joints and for joints not easily accessible clinically
(ie, subtalar and tarsal joints). We injected triamcinolone
hexacetonide at a dose of 1 mg/kg (maximum 40 mg) in
knees and shoulders, 0·75 mg/kg (maximum 30 mg) in
ankles and elbows, and 0·25–0·5 mg/kg (maximum
20 mg) in wrists. The dose of methylprednisolone acetate
was 5–10 mg for small hand and foot joints and 20–40 mg
www.thelancet.com Vol 389 March 4, 2017 911
Articles
for subtalar and intertarsal joints. We gave methotrexate
orally at a dose of 15 mg/m² (maximum 20 mg) once a
week, plus folinic acid (25–50% of the methotrexate dose
in mg, the day after methotrexate administration).
Methotrexate was started in all patients within 1 week of
joint injection. No injection was done under ultrasound
guidance. Most younger patients (at investigator
discretion) or patients who underwent the intra-articular
corticosteroids joints in two or more joints had the
procedure carried out under general anaesthesia. A non-
weight bearing period of 24 h was prescribed after
injection of corticosteroids.15 We did clinical assessments
at 1, 3, 6, and 12 months after treatment start. Personnel
at the coordinating centre at Gaslini Institute in Genoa,
Italy, verified the calculations of arthritis remission and
rate of flare. The study database was locked after the last
randomised patient had completed the 12 month follow-
up phase.
Outcomes
The primary outcome was the proportion of patients at
12 months who had remission of arthritis, determined by
clinician assessment and defined as the absence of any
clinical sign of active arthritis (swelling or, if swelling
was not present, pain and restricted motion) in all
injected joints. Prespecified secondary outcomes were
the proportion of patients who had relapse of synovitis
(flare) in specific joints (defined as recurrence of clinical
signs of active arthritis); the time to flare (defined as the
interval between the study start and the recurrence of
clinical signs of active arthritis); the proportion of
patients who achieved inactive disease, defined according
to Wallace and colleagues24 and clinical Juvenile Arthritis
Disease Activity Score (cJADAS)25 criteria, at 12 months;
frequency of intra-articular corticosteroids procedures
with or without general anaesthesia over 12 months; and
safety. We assessed safety by recording adverse events
and serious adverse events, which were re-coded with the
Medical Dictionary for Regulatory Activities classification
by system organ class and preferred term. As an
additional exploratory analysis, we compared the
proportions of patients between the two therapeutic
groups who developed new-onset uveitis.
Statistical analysis
Our original sample size calculation was based on our
previous analysis of intra-articular corticosteroid therapy
in 94 children with juvenile idiopathic arthritis.
Assuming a withdrawal rate of 5%, we originally
calculated that 252 people would be needed to show
significance, at 80% power with an alpha of 0·05, if
methotrexate increased remission rates by 30%. We
modified this prespecified sample size to base it on more
recent data with a much larger patient sample (n=440),18
and received approval by the ethics committee of the
coordinating centre on Oct 23, 2012. We calculated that a
sample size of 103 patients would be needed in each
 Articles
226 patients screened
207 randomised
102 randomly assigned
to intra-articular
corticosteroids
2 classified as having treatment
failure
2 lost due to change of address
or phone number
100 completed therapy
102 included in
intention-to-treat
analysis
Figure 1: Trial profile
For the clinical study protocol
see https://www.printo.it/docs/
WI_%20AIFA_ENGL.pdf
912 www.thelancet.com Vol 389 March 4, 2017
19 did not consent to be included
105 randomly assigned
to intra-articular
corticosteroids plus
methotrexate
13 classified as having treatment
failure
1 stopped methotrexate due
to an adverse event
1 withdrew consent
2 change of treatment due
to iridocyclitis
6 non-adherence to treatment
3 lost due to change of address
or phone number
92 completed therapy
105 included in
intention-to-treat
analysis
study group (total 206 patients) to have 80% power to
show that the addition of methotrexate increased the
remission rate from 55% to 75%.
We summarised baseline characteristics and efficacy and
safety variables with descriptive statistics. To assess
proportions, we used the χ² test or Fisher’s exact test, as
appropriate; for continuous variables we used the Mann-
Whitney U test. We used the Kaplan-Meier method to
produce survival curves for time to flare and compared
them with the log-rank test. We judged a p value of less
than 0·05 to be significant.
The primary and secondary outcomes were assessed by
intention to treat, with analyses including all randomised
patients who started treatment. The protocol specified that
patients who had new onset of arthritis in initially
unaffected joints, leading the physician to give another
intra-articular corticosteroid injection or start systemic
medication, should be regarded as having treatment
failure. We made this decision because we were concerned
that this new therapeutic intervention could affect (ie,
enhance) the probability of achieving the primary outcome
in injected joints. Patients who withdrew early for other
reasons (eg, because of consent withdrawal after receiving
treatment assignment, non-adherence to study, systemic
treatment change due to new onset or flare of iridocyclitis,
or occurrence of an adverse event) were also regarded as
having treatment failure from study withdrawal onwards.
Topical eye therapy for new-onset uveitis did not lead to
patient exclusion from the study.
The cumulative probability of remission of arthritis in all
injected joints at 6 and 12 months was computed through
the Kaplan-Meier approach, as follows. At time 0 (ie, study
baseline) all participants were considered at risk and the
probability of survival (ie, remission) (St0) was 1 (or 100%).
The probability of remission over time was computed
using the formula:
St+1 = St*((Nt+1–Dt+1)/Nt+1)
where N is the number of subjects at risk at study
baseline, t is the time and D is the number of subjects who
have the event of interest (ie, disease flare).
Because the loss to follow-up was high and varied
substantially between treatment arms, we imputed
missing covariate and outcome data by 20 imputations
with chained equations.26 Missing data imputation was
done in response to peer review and was not pre-
specified. The imputed data were analysed in a logistic
regression model with the primary outcome at 12 months
as dependent variable. All demographic and clinical
characteristics were eligible for inclusion in the model to
search for potential predictors of treatment response.
Evidence for interaction between the treatment effect
and the clinical variables was explored by adding the
first-order interaction term to the model. Significant
differences between the model with and without the
interaction term were assessed using the likelihood-ratio
test. Continuous variables were not dichotomised. The
final model included the treatment effect and all clinical
variables that showed an association with the outcome.
To account for differences in the occurrence of arthritis
flares over time, a Cox proportional hazards model was
applied to analyse the first episode of flare in injected
joints. Patients considered as having treatment failure for
the logistic regression model were also considered as
having treatment failure for the Cox model at the time of
the occurrence of flare. Patients with sustained remission
were censored at the time of the last observation. Baseline
variables were entered also in this model to find potential
predictors of treatment response.
The statistician who performed prespecified statistical
analyses (AP) was blinded to the aspects of the protocol
and results that did not involve statistical evaluations. We
acknowledge, however, that the two other investigators
(AC and EHPvD) who performed most statistical analyses
requested by the reviewers of the manuscript were
unblinded to the non-statistical aspects of the study. We
used Statistica version 9.1 for descriptive and univariate
analyses, Stata version 11.0 for calculation of binomial
exact confidence intervals, and R statistics version 3.0.0 for
multiple imputation, logistic regression, and survival
analysis. This study is registered with the European Union
Clinical Trials Register, EudraCT number 2008-006741-70.
 Articles

Role of the funding source

Intra-articular Intra-articular
The funder had no role in study design, data collection, data corticosteroids corticosteroids plus
analysis, data interpretation, or writing of the report, but did alone group methotrexate group
review the final report before submission. The corresponding (n=102) (n=105)
author had full access to all data in the study and had final Girls 73 (72%) 84 (80%)
responsibility for the decision to submit for publication. Boys 29 (28%) 21 (20%)
Age at onset (years) 2·8 (1·6–6·0) 2·5 (1·7–4·7)
Results Age at baseline (years) 4·5 (2·2–10·4) 4·1 (2·4–8·9)
Between July 7, 2009 and March 31, 2013, we screened Disease duration (months) 6·7 (2·8–19·5) 6·9 (3–22·0)
226 participants and randomly assigned 207 (figure 1). Positive antinuclear antibodies 73/87 (84%) 73/86 (85%)
The last patient completed the study on March 31, 2014. Previous use of prednisone 2 (2%) 4 (4%)
Participants had the typical features of children with Previous intra-articular 45 (44%) 46 (44%)
oligoarthritis: young age at disease onset, marked female corticosteroid injection
prevalence, and high frequency of antinuclear antibody Concurrent use of non-steroidal 9 (9%) 12 (11%)
positivity (table 1). The study population had early disease anti-inflammatory drugs
(table 1). There were no imbalances between treatment Physician’s global assessment 5 (4·0–6·3) 5 (4·0–6·0)
groups, apart from higher erythrocyte sedimentation rate Erythrocyte sedimentation rate 22 (9–40) 32 (15–54)
(mm/h)
in the group assigned to intra-articular corticosteroids
C-reactive protein (mg/dl) 0·5 (0·5–1·5) 0·5 (0·5–1·4)
plus methotrexate (p=0·0382).
48 (23%) patients were injected with corticosteroids in Number of injected joints 2 (2–3) 2 (2–3)

one joint and 159 (77%) patients in two or more joints. A Patients with 1 injected joint 22 (22%) 26 (25%)

total of 490 joints were injected, most frequently the knee Patients with >1 injected joints 80 (78%) 79 (75%)
and the ankle (table 2). Of these, 84 (20%) joints were Table 1: Baseline demographic and disease characteristics
injected with methylprednisolone acetate, with similar
proportions between treatment groups (19% vs 21%). The
elbow and the wrist were more commonly injected in the Intra-articular Intra-articular corticosteroids plus
methotrexate therapy group (table 2). Frequency of intra- corticosteroids alone group methotrexate group
(n=102) (n=105)
articular corticosteroids procedures with or without
general anaesthesia is not reported. Patients, n (%) Joints, n Patients, n (%) Joints, n p value*
In the intention-to-treat analysis, 35 (34%) patients Shoulder 1 (1%) 2 1 (1%) 1 0·99
assigned to intra-articular corticosteroids alone and Elbow 2 (2%) 2 10 (10%) 12 0·0199
41 (39%) assigned to intra-articular corticosteroids plus Wrist 1 (1%) 1 8 (8%) 9 0·0352
methotrexate had remission of arthritis in all injected Metacarpophalangeal 10 (10%) 10 5 (5%) 5 0·16
joints at 12 months (p=0·48). 18 (18%) patients in the Proximal interphalangeal 11 (11%) 12 15 (14%) 15 0·45
intra-articular corticosteroids alone group and 21 (20%) Distal interphalangeal 2 (2%) 2 0 ·· 0·24
patients in the intra-articular corticosteroids plus Knee 80 (78%) 113 74 (70%) 106 0·19
methotrexate groups had remission of arthritis in all Ankle 46 (45%) 54 58 (55%) 70 0·14
injected joints but had new onset of arthritis in previously Subtalar 23 (22·5) 25 25 (24%) 29 0·14
unaffected joints before 12 months and so were not Tarsal 0 ·· 1 (1%) 1 0·99
included in the intention-to-treat analysis. When including Metatarsophalangeal or foot 1 (1%) 1 4 (4%) 4 0·37
those classed as having treatment failure in the intention- interphalangeal
to-treat analysis, the total number of patients who achieved Total number of injected joints ·· 228 ·· 262 ··
remission in injected joints was 53 (52%) assigned to
Data are median (IQR), n (%), or n/N (%). No patient had previously received methotrexate or other synthetic or
intra-articular corticosteroids alone and 62 (59%) to intra-
biological disease modifying anti-rheumatic drugs. *For the comparison of the proportion of patients who had each joint
articular corticosteroids plus methotrexate (p=0·3050). injected between the two groups.
New-onset arthritis occurred in 38 (37%) patients
Table 2: Joints injected
allocated to intra-articular corticosteroids alone and in
38 (36%) allocated to intra-articular corticosteroids plus
methotrexate. The proportion of patients with inactive alone (49%; 95% CI 39–58; and 35%, 25–44; respectively)
disease at 12 months was 26·5% (n=27) in the intra- than in children assigned intra-articular injection plus
articular corticosteroids alone group and 36% (n=38) in the methotrexate (67%, 56–75; and 46%, 35–56; respectively).
intra-articular corticosteroids plus methotrexate group In the group assigned to intra-articular corticosteroids
(p=0·13) by the Wallace criteria,24 and 27·5% (n=28) and alone, injected joints in which flare occurred most
28% (n=29), respectively (p=0·98), by the cJADAS criteria.25 frequently were the following: metacarpophalangeal (six
There was a lower cumulative probability of remission [60%] joints), elbow (one [50%] of two), subtalar (12 [48%]
of arthritis in all injected joints at 6 months and of 25), and ankle (22 [41%] of 54. Lack of arthritis
12 months in children allocated intra-articular injection remission was seen in 23 (20%) of 113 injected knees.

www.thelancet.com Vol 389 March 4, 2017 913

 Articles
100 Intra-articular corticosteroids group
Intra-articular corticosteroids
and methotrexate group
75
Flare-free survival (%)
50
25
HR 0·67 (95% CI 0·46–0·97); log-rank p=0·0321
0
0 2 4 6 8 10 12
Number at risk
Time from enrolment (months)
Intra-articular 102 85 63 52 40 38 25
corticosteroids group
Intra-articular 105 92 71 64 53 42 26
corticosteroids and
methotrexate group
Figure 2: Time to flare of arthritis in injected joints
In the group assigned to intra-articular corticosteroids
plus methotrexate, injected joints in which flare occurred
most frequently were the following: subtalar (12 [41%]
of 29), ankle (18 [26%] of 70), elbow (three [25%] of 12),
and metacarpophalangeal (1 [10%] of 10). Lack of arthritis
remission was seen in 13 (12%) of 106 of injected knees.
Time to arthritis flare in injected joints was longer for
patients assigned to intra-articular corticosteroids plus
methotrexate (median 10·1 months, 95% CI 7·6 to >16)
than for those assigned to intra-articular corticosteroids
alone (6·0 months, 4·6–8·2; figure 2).
The analysis of imputed covariate and outcome data in
a univariable logistic-regression model showed no effect
for the addition of methotrexate (odds ratio [OR] 0·69,
95% CI 0·38–1·24; p=0·22). However, multivariable
analysis that included treatment effect and erythrocyte
sedimentation rate (included as continuous variable)
showed that concomitant administration of methotrexate
was protective against the occurrence of arthritis flare,
although the statistical effect was marginal (adjusted
OR 0·53, 95% CI 0·27–1·01; p=0·05). Furthermore,
higher erythrocyte sedimentation rate predicted arthritis
flare (adjusted OR 1·03, 95% CI 1·01–1·05; p=0·007).
However, there was no evidence of interaction between
erythrocyte sedimentation rate and treatment effect
(p=0·28 in the likelihood-ratio test).
In the Cox proportional hazards model, the hazard ratio
(HR) of flare of arthritis in injected joints for intra-articular
corticosteroids plus methotrexate versus intra-articular
corticosteroids alone in univariable analysis was 0·67
(95% CI 0·46–0·97, p=0·0321). In multivariable models,
the HR adjusted for the erythrocyte sedimentation rate
(included as continuous variable) was 0·55 (95% CI
0·37–0·81, p=0·003). A higher erythrocyte sedimentation
rate was associated with a greater risk of arthritis flare
914 www.thelancet.com Vol 389 March 4, 2017
(HR 1·02, 95% CI 1·01–1·02; p=0·0002). The same
analyses performed without imputed data yielded similar
results (data not shown).
Among the 171 patients for whom the information was
available, three (4%) patients assigned to intra-articular
corticosteroids alone and six (8%) to intra-articular
corticosteroids plus methotrexate had new-onset uveitis
(p=0·4957). Systemic adverse events were recorded for
20 (17%) patients who received methotrexate and
included gastrointestinal discomfort (nausea, vomiting,
or constipation; n=14), increased liver transaminases
(n=8), fatigue (n=2), irritability (n=2), hair loss (n=1),
and leukopenia (n=1). Increased liver transaminases led
14 16 to permanent treatment discontinuation in one patient,
temporary drug discontinuation in three patients, and
0 0 dose reduction in two patients. Gastrointestinal
4 1
discomfort led to permanent treatment discontinuation
in one patient. No patient had a serious adverse event.
No patient developed skin hypopigmentation,
subcutaneous atrophy, or other side-effects at the site of
intra-articular injection. No side-effects related to
systemic absorption of intra-articular corticosteroids
were recorded.
Discussion
This is the first randomised controlled trial designed to
test the superiority of intra-articular corticosteroids
combined with systemically administered therapy with
methotrexate compared with intra-articular cortico­
steroids alone in oligoarticular juvenile idiopathic
arthritis. Patients in this study had, on average, a short
disease duration, which suggests that our analysis
provides insights into the effectiveness of the two
interventions used as early treatment strategies. We
chose a 12 month follow-up time because we previously
found that the first year after the injection could
constitute a therapeutic window, in which the
combination of local injection with a systemic
intervention might help to maximise the long-term
benefit.18,22 Children will be followed-up for up to 2 years
for the evaluation of long-term safety profile.
Our findings suggest that concomitant administration
of oral methotrexate might prolong and, to a lesser
extent, increase the effectiveness of intra-articular
corticosteroid therapy in children with oligoarticular
juvenile idiopathic arthritis. Although there was no
difference between the two therapeutic groups in the
achievement of the primary outcome at 12 months in the
intention-to-treat analysis, the analysis of the imputed
covariate and outcome data with multivariable logistic
regression and Cox proportional hazards model favoured
the combination of intra-articular corticosteroid therapy
with methotrexate over intra-articular corticosteroid
therapy alone. Although higher erythrocyte sedi­
mentation rate predicted arthritis flare, there was no
evidence of interaction between erythrocyte sedi­
mentation rate and treatment effect. Time to arthritis

flare analyses also favoured the combination of intra-
articular corticosteroids with methotrexate over treatment
with intra-articular corticosteroids alone. The intention-
to-treat analysis findings might be of less value than our
other analyses for two reasons: first, it did not account for
the high proportion of patients who had remission of
arthritis in all injected joints, but were considered as
having treatment failure because of new onset of arthritis
in previously unaffected joints before the 12 month
evaluation; second, it could underestimate the effect of
methotrexate owing to the greater number of early
withdrawals in the combination therapy group than in
the group treated with intra-articular corticosteroids
alone (13 vs two); and third, it did not correct for the effect
of the erythrocyte sedimentation rate.
Although our results in the intention-to-treat population
were not statistically significant, we feel that these findings
lend support to those of previous non-controlled studies22,23
and indicate that the outcome of intra-articular
corticosteroid therapy might be improved if this procedure
was combined with the administration of methotrexate.
This recommendation does not generalise to children
with monoarthritis of the knee who are candidates to their
first intra-articular injection because these patients were
excluded from our study. It is our policy to treat all these
patients initially with local injection therapy only. In case
of short-term relapse of arthritis in the injected knee or
extension of arthritis to other joints in the first
6–12 months, methotrexate is added.
The finding that a sizeable proportion of children
developed arthritis in initially unaffected joints was
expected, given the distinctive propensity of oligoarticular-
onset juvenile idiopathic arthritis to spread over time
and the short median disease duration noted in our
population. In a previous analysis of children with
antinuclear antibodies (ANA)-positive juvenile idiopathic
arthritis (who constituted the large majority of the study
sample; table 1), we found that, at disease presentation,
72% of patients had involvement of a single joint
(monoarthritis), 23% had involvement of two to four
joints (oligoarthritis), and only 5% had involvement of
five or more affected joints (polyarthritis). After 6 months,
the frequency of monoarthritis was decreased to 33%,
whereas that of oligoarthritis and polyarthritis was
increased to 48% and 18%, respectively.27 It is surprising
that in our study new arthritis in new joints occurred
with equal frequency in both treatment groups
considering that methotrexate is a systemic medication
shown to be effective in controlling polyarticular juvenile
idiopathic arthritis.19,20 This observation suggests that oral
methotrexate might not prevent arthritis extension in
children with oligoarticular-onset juvenile idiopathic
arthritis. Whether subcutaneous methotrexate, systemic
corticosteroids, or biologic drugs have a better preventive
capacity, remains to be established.
We chose oral methotrexate because we felt that
subcutaneous injections, which can cause discomfort to
www.thelancet.com Vol 389 March 4, 2017 915
Articles
children, were not justified in a disease subset that is
considered the most benign and in which methotrexate
is not universally used as first-line therapy. A recent
retrospective study demonstrated no differences in
effectiveness between the routes of administration
among children who remained on methotrexate
monotherapy.28 However, we cannot exclude a chance
that the parenteral route could have been of greater
benefit. Other paediatric studies and the experience in
adults with rheumatoid arthritis have shown increased
bioavailability of subcutaneous compared to oral
methotrexate at higher doses,29 as well as improvement
upon switching from oral to subcutaneous
administration.30
Our study has a number of limitations. First, masking
was not implemented in our trial for ethical, logistical
(double-dummy design), and economic reasons and
assessors at each participating centre were not masked to
assignments, which might have led to some bias. Blinded
assessment was not feasible in some participating
centres due to the availability of only a single paediatric
rheumatologist. However, the primary outcome measure
used was simple and straightforward for any experienced
paediatric rheumatologist. Second, the use of
methylprednisolone acetate instead of triamcinolone
hexacetonide for injections in small and difficult-to-
access joints might have reduced the effectiveness of
injections because this corticosteroid preparation is less
efficacious. However, the proportion of joints injected
with methylprednisolone acetate was similar in the two
study groups. The administration of a more soluble
corticosteroid molecule for the most challenging
procedures may explain the absence of subcutaneous
atrophy, which is a well-known side-effect of
triamcinolone hexacetonide.15 Notably, triamcinolone
hexacetonide is no longer available in the USA. Third, we
did not use ultrasound guidance, which might have
increased the precision of the procedure and, potentially,
its effectiveness. However, its application would require a
practitioner who is familiar with the technique. Finally,
no system was implemented to guarantee that
methotrexate was given to the children to protocol.
In conclusion, our study suggests that concomitant
administration of methotrexate might prolong and, to a
lesser extent, augment the effectiveness of intra-articular
corticosteroid therapy in children with oligoarticular
juvenile idiopathic arthritis, without an appreciable
increase in serious toxicity. This combination could be
considered as reference treatment in everyday clinical
practice for paediatricians, particularly in children with
higher erythrocyte sedimentation rate. Furthermore,
owing to its high remitting potential, it could gain a
central role in treat to target strategies for children with
chronic arthritis. These indications should be balanced
against the risk of methotrexate side-effects. The addition
of methotrexate did not reduce the prevalence of new-
onset of disease in previously unaffected joints,
 Articles
underscoring the need for future clinical trials to
scrutinise the capability of therapeutic interventions to
prevent arthritis extension in children with oligoarticular-
onset juvenile idiopathic arthritis.
Contributors
AR, AP, NR, and AM jointly designed the study. Data were collected by
members of the Italian Paediatric Rheumatology Study Group. AP, AC,
and EHPvD did the statistical analysis . The first and subsequent
versions of the report were written by AR and AC, edited by EHPvD, NR
and AM, and revised critically by all authors. All authors participated in
data collection, data analysis, and data interpretation, and have approved
the final study report.
Declaration of interests
AR reports personal fees from AbbVie, Bristol-Myers Squibb, Novartis,
Pfizer, Roche, and Johnson & Johnson. AP reports personal fees from
Boehringer, Omniprex, and Novartis. FDB reports grants from
Hoffmann-La Roche, Bristol-Myers Squibb, Novimmune, Novartis, Abbot,
Pfizer, and Sobi. NR reports personal fees from AbbVie, Amgen,
Biogenidec, Alter, AstraZeneca, Baxalta Biosimilars, Biogenidec,
Boehringer, Bristol-Myers Squibb, Celgene, CrescendoBio, EMD Serono,
Hoffman-La Roche, Italfarmaco, Janssen, MedImmune, Medac, Novartis,
Novo Nordisk, Pfizer, Sanofi-Aventis, Servier, Takeda, and UCB
Biosciences; and other financial relationships from Bristol-Myers Squibb,
GlaxoSmithKline, Hoffman-La Roche, Novartis, Pfizer, Sanofi-Aventis,
Schwarz Biosciences, Abbott, Francesco Angelini SPA, Sobi, and Merck
Serono. AM reports personal fees from Abbvie, Boehringer, Celgene,
CrescendoBio, Janssen, Meddimune, Novartis, NovoNordisk, Pfizer,
Sanofi-Aventis, Vertex, and Servier; and other financial relationships from
BMS, GlaxoSmithKline, Hoffman-La Roche, Novartis, Pfizer,
Sanofi-Aventis, Schwarz Biosciences, Abbott, Francesco Angelini SPA,
Sobi, and Merck Serono. All other authors declare no competing interests.
Acknowledgments
This study was funded by a grant from the Italian Agency of Drug
Evaluation (to AR) and was supported by the Istituto Giannina Gaslini,
Genoa, Italy.
References
1 Petty RE, Southwood TR, Manners P, et al. International League of
Associations for Rheumatology classification of juvenile idiopathic
arthritis: second revision, Edmonton, 2001. J Rheumatol 2004;
31: 390–92.
2 Ravelli A, Martini A. Juvenile idiopathic arthritis. Lancet 2007;
369: 767–78.
3 Wallace CA, Huang B, Bandeira M, Ravelli A, Giannini EH.
Patterns of clinical remission in select categories of juvenile
idiopathic arthritis. Arthritis Rheum 2005; 52: 3554–62.
4 Vidqvist KL, Malin M, Varjolahti-Lehtinen T, Korpela MM.
Disease activity of idiopathic juvenile arthritis continues through
adolescence despite the use of biologic therapies. Rheumatology
2013; 52: 1999–2003.
5 Consolaro A, Ravelli A. Unraveling the phenotypic variability of
juvenile idiopathic arthritis across races or geographic
areas—key to understanding etiology and genetic factors?
J Rheumatol 2016; 43: 683–85.
6 Guillaume S, Prieur AM, Coste J, Job-Deslandre C.
Long-term outcome and prognosis in oligoarticular-onset juvenile
idiopathic arthritis. Arthritis Rheum 2000; 43: 1858–65.
7 Oen K, Malleson PN, Cabral DA, Rosenberg AM, Petty RE,
Cheang M. Disease course and outcome of juvenile rheumatoid
arthritis in a multicenter cohort. J Rheumatol 2002; 29: 1989–99.
8 Lovell DJ, Ruperto N, Giannini EH, Martini A. Advances from clinical
trials in juvenile idiopathic arthritis. Nat Rev Rheumatol 2013; 9: 557–63.
9 Ruperto N, Nikishina I, Pachanov ED, et al. A randomized,
double-blind clinical trial of two doses of meloxicam compared with
naproxen in children with juvenile idiopathic arthritis: Short-and
long-term efficacy and safety results. Arthritis Rheum 2005; 52: 563–72.
10 Reiff A, Lovell DJ, Adelsberg JV, et al. Evaluation of the comparative
efficacy and tolerability of rofecoxib and naproxen in children and
adolescents with juvenile rheumatoid arthritis: a 12-week
randomized controlled clinical trial with a 52-week open-label
extension. J Rheumatol 2006; 33: 985–95.
916 www.thelancet.com Vol 389 March 4, 2017
11 Sobel RE, Lovell DJ, Brunner HI, et al. Safety of celecoxib and
nonselective nonsteroidal anti-inflammatory drugs in juvenile
idiopathic arthritis: results of the Phase 4 registry.
Pediatr Rheumatol Online J 2014; 12: 29.
12 Zulian F, Martini G, Gobber D, Plebani M, Zacchello F, Manners P.
Triamcinolone acetonide and hexacetonide intra-articular treatment
of symmetrical joints in juvenile idiopathic arthritis: a double-blind
trial. Rheumatology 2004; 43: 1288–91.
13 Beukelman T, Guevara JP, Albert DA, Sherry DD, Burnham JM.
Variation in the initial treatment of knee monoarthritis in juvenile
idiopathic arthritis: a survey of pediatric rheumatologists in the
United States and Canada. J Rheumatol 2007; 34: 1918–24.
14 Beukelman T, Patkar NM, Saag KG, et al. 2011 American College of
Rheumatology recommendations for the treatment of juvenile
idiopathic arthritis: Initiation and safety monitoring of therapeutic
agents for the treatment of arthritis and systemic features.
Arthritis Care Res 2011; 63: 465–82.
15 Scott C, Meiorin S, Filocamo G, et al. A reappraisal of intra-articular
corticosteroid therapy in juvenile idiopathic arthritis.
Clin Exp Rheumatol 2010; 28: 774–81.
16 Bloom BJ, Alario AJ, Miller LC. Intra-articular corticosteroid therapy
for juvenile idiopathic arthritis: report of an experiential cohort and
literature review. Rheumatol Int 2011; 31: 749–56.
17 Sherry DD, Stein LD, Reed AM, Schanberg LE, Kredich DW.
Prevention of leg length discrepancy in young children with
pauciarticular juvenile rheumatoid arthritis by treatment with
intraarticular steroids. Arthritis Rheum 1999; 42: 2330–34.
18 Lanni S, Bertamino M, Consolaro A, et al. Outcome and predicting
factors of single and multiple intra-articular corticosteroid
injections in children with juvenile idiopathic arthritis.
Rheumatology 2011; 50: 1627–34.
19 Giannini EH, Brewer EJ, Kuzmina N, et al. Methotrexate in resistant
juvenile rheumatoid arthritis. Results of the U.S.A.-U.S.S.R.
double-blind, placebo-controlled trial. N Engl J Med 1992; 326: 1043–49.
20 Ruperto N, Murray KJ, Gerloni V, et al. A randomized trial of parenteral
methotrexate comparing an intermediate dose with a higher dose in
children with juvenile idiopathic arthritis who failed to respond to
standard doses of methotrexate. Arthritis Rheum 2004; 50: 2191–201.
21 Papadopoulou C, Kostik M, Bohm M, et al. Methotrexate therapy
may prevent the onset of uveitis in juvenile idiopathic arthritis.
J Pediatr 2013; 163: 879–84.
22 Papadopoulou C, Kostik M, Gonzalez-Fernandez MI, et al.
Delineating the role of multiple intraarticular corticosteroid
injections in the management of juvenile idiopathic arthritis in the
biologic era. Arthritis Care Res 2013; 65: 1112–20.
23 Marti P, Molinari L, Bolt IB, Seger R, Saurenmann RK.
Factors influencing the efficacy of intra-articular steroid injections
in patients with juvenile idiopathic arthritis. Eur J Pediatr 2008;
167: 425–30.
24 Wallace CA, Giannini EH, Huang B, Itert L, Ruperto N.
American College of Rheumatology provisional criteria for defining
clinical inactive disease in select categories of juvenile idiopathic
arthritis. Arthritis Care Res 2011; 63: 929–36.
25 Consolaro A, Negro G, Chiara Gallo M, et al. Defining criteria for
disease activity states in nonsystemic juvenile idiopathic arthritis
based on a three-variable juvenile arthritis disease activity score.
Arthritis Care Res 2014; 66: 1703–09.
26 Van Buuren S, Groothuis-Oudshoorn K. Mice: Multivariate
Imputation by Chained Equations in R. J Stat Softw 2011; 45: 1–67.
27 Felici E, Novarini C, Magni-Manzoni S, et al. Course of joint disease
in patients with antinuclear antibody-positive juvenile idiopathic
arthritis. J Rheumatol 2005; 32: 1805–10.
28 Klein A, Kaul I, Foeldvari I, Ganser G, Urban A, Horneff G. Efficacy
and safety of oral and parenteral methotrexate therapy in children
with juvenile idiopathic arthritis: an observational study with
patients from the German Methotrexate Registry. Arthritis Care Res
2012; 64: 1349–56.
29 Bianchi G, Caporali R, Todoerti M, Mattana P. Methotrexate and
Rheumatoid Arthritis: Current Evidence Regarding Subcutaneous
Versus Oral Routes of Administration. Adv Ther 2016; 33: 369–78.
30 Alsufyani K, Ortiz-Alvarez O, Cabral DA, Tucker LB, Petty RE,
Malleson PN. The role of subcutaneous administration of
methotrexate in children with juvenile idiopathic arthritis who have
failed oral methotrexate. J Rheumatol 2004; 31: 179–82.