Co

ORIGINAL ARTICLE: GASTROENTEROLOGY

Adalimumab Therapy in Children With Crohn Disease

Previously Treated With Infliximab

�Martinus Cozijnsen, yVera Duif, zFreddy Kokke, §Angelika Kindermann, jjPatrick van Rheenen,�Tim de Meij, #Maaike Schaart, yGerard Damen, ��Obbe Norbruis, yyRolf Pelleboer,

zzAnita Van den Neucker, §§Herbert van Wering, jjjjThalia Hummel, ��Johanna Oudshoorn,�Johanna Escher, and �Lissy de Ridder, on Behalf of the Dutch PIBD Working

pyright 2015 by

Group Kids with Crohn and Colitis

by 50% for 2 years. E

remission (decrease in w

Received March 31, 2014From �Pediatric Gastr

Children’s Hospital, RUniversity Nijmegenterology, WilhelminaUtrecht, §PediatricAcademic Medical CBeatrix Children’s HGroningen, �PediatrCenter, Amsterdam,Kinderziekenhuis-Lei��Pediatric GastroeGastroenterology, Caenterology, Maastric§§Pediatric GastroenGastroenterology, M��Pediatric Gastroentlands.

Address correspondence aDepartment of PediChildren’s Hospital,2060, 3000 CB Rottasmusmc.nl).

The authors report no coCopyright # 2015 by E

Hepatology, and NutGastroenterology, He

DOI: 10.1097/MPG.0000

JPGN � Volume 60, N

ABSTRACT

Objectives: Adalimumab, a humanised anti-tumour necrosis factor anti-

body, is an effective treatment in adult patients with refractory Crohn disease

(CD). The available literature on its efficacy in children remains limited. We

aimed to evaluate the real-world efficacy in paediatric patients with CD and

compare the efficacy between infliximab (IFX) nonresponders and patients

who lost response to IFX.

Methods: All Dutch patients with CD receiving adalimumab before the age

of 18 years after previous IFX therapy were identified. We analysed

longitudinal disease activity, assessed by the mathematically weighted

Pediatric Crohn’s Disease Activity Index (wPCDAI) or the physician

global assessment (PGA), and adverse events (AEs).

Results: Fifty-three patients with CD were included. Twelve patients

received monotherapy and the others received combination treatment

with thiopurines (n¼ 21), methotrexate (n¼ 11), steroids (n¼ 7), or

exclusive enteral nutrition (n¼ 2). Median follow-up was 12 months

(interquartile range 5–23). Remission was reached in 34 patients (64%,

wPCDAI< 12.5 or PGA¼ 0) after a median of 3.3 months, and maintained

ESPGHAN and NASPGHAN. Un

leven patients (21%) reached response but not

PCDAI� 17.5 or decrease in PGA). Eighteen

; accepted September 27, 2014.oenterology, Erasmus Medical Center-Sophiaotterdam, yPediatric Gastroenterology, Radboud

Medical Centre, Nijmegen, zPediatric Gastroen-Children’s Hospital-University Medical Center,

Gastroenterology, Emma Children’s Hospital-enter, Amsterdam, jjPediatric Gastroenterology,ospital-University Medical Center Groningen,

ic Gastroenterology, VU University Medical#Pediatric Gastroenterology, Willem-Alexanderden University Medical Center, Leiden,nterology, Isala Hospital, Zwolle, yyPediatrictharina Hospital, Eindhoven, zzPediatric Gastro-ht University Medical Center, Maastricht,

terology, Amphia Hospital, Breda, jjjjPediatricedisch Spectrum Twente, Enschede, and

erology, Gelre Hospitals, Apeldoorn, The Nether-

nd reprint requests to Lissy de Ridder, MD, PhD,atric Gastroenterology, Erasmus MC—SophiaUniversity Medical Center Rotterdam, PO Boxerdam, The Netherlands (e-mail: l.deridder@er-

nflicts of interest.uropean Society for Pediatric Gastroenterology,rition and North American Society for Pediatricpatology, and Nutrition000000000589

umber 2, February 2015

patients (34%) failed adalimumab treatment because of nonresponse

(n¼ 4), lost response (n¼ 11), or AEs (n¼ 3). More IFX nonresponders

failed adalimumab treatment than patients who lost response to IFX (2/3 vs

8/34, hazard ratio 18.8, 95% confidence interval 1.1–303.6). Only 1 patient

encountered a serious AE, a severe but nonfatal infection.

Conclusions: In clinical practice, adalimumab induces remission in two-

thirds of children with IFX refractory CD.

Key Words: adalimumab, biological, Crohn disease, infliximab, paediatric

(JPGN 2015;60: 205–210)

C rohn disease (CD) is a chronic inflammatory bowel disease(IBD) that presents before the age of 20 years in 25% to 30%

of patients (1). In the Netherlands, the reported incidence of CD inchildren and adolescents younger than the age of 18 is 2.1/100,000(2). CD manifests more severely in childhood and adolescence thanin adulthood, with growth retardation and delayed puberty, worsemalnutrition, and a higher risk of complications such as strictures,abscesses, or fistulas (1,3).

For paediatric patients with CD, first-line treatment consistsof exclusive enteral nutrition or corticosteroids to induce remission,and thiopurines or methotrexate to maintain remission. If this first-line of treatment fails, anti-tumour necrosis factor (TNF) antibodytherapy is often indicated (4).

The first anti-TNF antibody that was studied in and regis-tered for patients with CD was infliximab (IFX), which is a chimeric(mouse/human) monoclonal antibody. IFX induces and maintainsremission in patients with CD (5,6). Unfortunately, one-third ofthese patients withdraw from IFX therapy within 3 years, and halfwithdrew within 5 years (6,7). Some withdraw because they neverresponded to the drug, but most because they lose response.

One known reason for loss of response to IFX is developmentof antibodies-to-infliximab (ATI) that neutralise the drug (8,9). Toreduce such an antibody reaction, adalimumab, a fully humanisedanti-TNF monoclonal antibody, was developed and tested for thetreatment of CD (10,11). The efficacy and safety in paediatricpatients have been studied in 1 large prospective study (12) and 2retrospective studies (13,14). The positive results from the pro-spective study have led to registration of adalimumab forpaediatric usage.

In this Dutch nationwide study, we aimed to evaluate thereal-world efficacy of adalimumab in paediatric patients with CDwho previously failed treatment with IFX. We included both

authorized reproduction of this article is prohibited.

patients who had no response to IFX and those who lost response,and compared adalimumab efficacy between these 2 groups.

205

Co

median duration of IFX treatment was 15.6 months (IQR 10.8–25.8). Reasons to stop with IFX were nonresponse (n¼ 3, 6%), loss

TABLE 1. Patient characteristics

n (%) or

median (IQR)

Total included patients 53

Sex Female 27 (50.9%)

Age, y At diagnosis 11 (8–13)

At adalimumab

commencement

14 (13–16)

Disease location�

Ileocecal (L1) 6 (11.3%)

Colonic (L2) 11 (20.8%)

Both ileal and

colonic (L3)

36 (67.9%)

Disease behaviour�

Not stricturing nor

penetrating (B1)

42 (79.2%)

Stricturing (B2) 8 (15.1%)

Penetrating (B3) 1 (1.9%)

Both penetrating and

stricturing (B4)

2 (3.8%)

History of perianal

disease

19 (35.8%)

Growth delay at diagnosis�

20 (37.7%)

Prior treatment Exclusive enteral

nutrition

32 (60.4%)

Steroids 50 (94.3%)

Mesalazine 27 (50.9%)

Thiopurines 52 (98.1%)

Methotrexate 26 (49.1%)

IFX treatment duration, mo 15.7 (10.8–25.8)

Type of IFX failure Lost response 34 (64.2%)

Allergic reaction 11 (20.8%)

Adverse effects 5 (9.4%)

Nonresponse 3 (5.7%)

ATI presence Tested 38 (71.7%)

Positive (>15 AE/mL) 21 (55.3%)

History of CD-related surgery 12 (22.6%)

METHODSWe performed a nationwide, observational cohort study and

invited all of the Dutch paediatric gastroenterologists, who pre-scribe biologicals to paediatric patients with IBD, to participate.Those who agreed were asked to identify all of the eligible patients.Inclusion criteria were as follows: diagnosed as having CD, treatedwith adalimumab before the age of 18, and treated with IFX beforethe start of adalimumab. Exclusion criteria were as follows: con-flicting comorbidity (such as autoimmune diseases or other chronicintestinal pathology), recorded bad treatment adherence, andprevious participation in the prospective study by Hyams et al(12). We collected the following information from patient records:patient characteristics, disease localization and behaviour, previoustreatment history, longitudinal disease activity from the start ofadalimumab treatment, and adverse events (AE).

Adalimumab Efficacy Assessment

To assess the efficacy of adalimumab therapy, diseaseactivity was retrospectively assessed by the first or second authorusing the mathematically weighted Pediatric Crohn’s DiseaseActivity Index (wPCDAI) (15) or a 4-stepped physician globalassessment (PGA) in case the wPCDAI could not be calculated.Disease activity was categorised into 4 categories: remission, mild,moderate, and severe.

Since timing of follow-up visits was unstructured, follow-upvisits were sought closest to and pooled at the following timepoints:1 month, 4 months, 8 months, and 12 months after the start oftreatment and then yearly depending on availability.

On the basis of the development in disease activity, the effect ofadalimumab at each time point was categorised as clinical response,clinical remission, no response, or loss of response. Clinical responsewas defined as a decrease in the wPCDAI by at least 17.5 points, or adecrease in PGA from either moderate or severe to mild. AwPCDAI< 12.5 or a PGA equalling zero was defined as remission.Loss of response was defined as an increased disease activity tomoderate or severe after having reached response or remission.

If adalimumab treatment was discontinued, we documentedthe reason for discontinuation. In the case of lack of efficacy(nonresponse, loss of response) or intolerance (allergic reactionor AEs), we considered these patients as having failed adalimumab.Secondly, we considered patients requiring CD-related surgeryduring follow-up as having failed adalimumab owing to loss ofefficacy. Follow-up of these patients stopped when surgery wasperformed, even when adalimumab therapy continued after.Patients requiring perianal surgery or surgery for reasons otherthan disease progression were not categorised as having failedadalimumab and follow-up in these patients therefore continued.

Statistical Analysis

Data collection and analysis were performed using SPSSversion 21 (IBM SPSS Statistics, Armonk, NY). Continuous non-parametric data are presented as median (interquartile range [IQR]);parametric data were absent. Categorical data are presented by thenumber of cases and the proportion of cases (%). Kaplan-Meieranalyses and Cox proportional hazard models were used to analyseand visualise adalimumab failure in relation to time. The Fisherexact test was used to compare categorical data between groups. Weused 0.05 as the cutoff point for statistical significance.

Ethical Considerations

de Ridder et al

pyright 2015 by ESPGHAN and NASPGHAN. Un

This study was exempted from institutional review boardapproval because it involved the collection of data generated by

206

routine medical care. The data were collected and recorded by theinvestigators in such a manner that subjects could not be identified,directly or through identifiers linked to the subjects.

RESULTSIn total 15 Dutch centres participated, that is, 10 centres of

the Dutch paediatric IBD working group (KiCC) and 5 additionalcentres. Between 2005 and 2013, 59 paediatric patients with CD hadswitched from IFX to adalimumab of whom 53 met the inclusionand exclusion criteria. Twelve of the 15 centres kept record of thenumber of patients with CD who received IFX during this period:351 received IFX of whom 55 switched to adalimumab (medianswitching rate 13.3% [IQR 9.8–18.1]). The patient characteristicsare displayed in Table 1.

Previous Treatment

Thirty-two patients (60%) had received at least 1 course ofexclusive enteral nutrition, 50 (94%) had been treated with steroids,27 (51%) with mesalazine, 52 (98%) with thiopurines (azathioprineand/or mercaptopurine), and 26 (49%) with methotrexate. The

JPGN � Volume 60, Number 2, February 2015

authorized reproduction of this article is prohibited.

ATI¼ antibodies to IFX; IFX¼ infliximab; IQR¼ interquartile range.�As defined by the Paris classification (20).

www.jpgn.org

Co

6 12 18 240

50

100

Rem

issi

on %

Follow-up mo

Months

Patients at risk

0

34

6

14

12

7

18

5

24

4

FIGURE 1. Duration of remission. Cumulative hazard curve displayingthe duration of remission, that is, the time from induction of remission

until a first relapse, for those patients reaching remission during

follow-up. After 6 months 70% of these patients maintained remission

Adalimumab Therapy in Children With IFX-Refractory CD

of response (n¼ 34, 64%), allergic reactions (n¼ 11, 21%), oradverse effects (n¼ 5, 9%).

In a large subgroup (n¼ 38, 72%) ATI had been tested beforeadalimumab commencement. Twenty-one of those had testedpositive (ATI concentration> 15 AE/mL), 16 of whom (76%)had lost response to IFX, 3 (14%) had developed an allergicreaction, and 2 (10%) had experienced adverse effects. The other17 patients had tested negative for ATI (2 with nonresponse, 11 wholost response, 3 with allergic reactions, and 1 with adverse effects).

Twelve patients (23%) had undergone partial or total bowelresection before adalimumab commencement, at a median age of14 years (IQR 11–14), a median of 21 months before the startof adalimumab therapy (IQR 11–33).

Adalimumab Treatment

Adalimumab induction regimens differed. Thirty-ninepatients (74%) started with a double dosage before the maintenancetreatment; the remainder received the maintenance dosage straightfrom the start. Initial maintenance doses were based on body weight(20–40 mg for patients <40 kg, 40–80 mg for patients >40 kg).Treatment escalation was needed in 13 patients (25%), performedby increasing the dose, shortening the dose interval, or both.

At the start of adalimumab 41 patients (77%) were usingconcomitant CD-related medication, including immunomodulators(thiopurines [n¼ 21] or methotrexate [n¼ 11]), steroids (n¼ 7), orexclusive enteral nutrition (n¼ 2).

Response to Adalimumab

We studied the children and teenagers in our cohort for a totalof 66 patient-years (PY), that is, for a median of 12 months perpatient (IQR 5–23). In 95% of the follow-up visits disease activitywas assessed by either the wPCDAI (45%) or the PGA (50%). Forthe remainder this was not possible. The timing of the follow-upvisits differed by a median of 14% (IQR 5–27) from thepresented timepoints.

At adalimumab commencement, 3 patients were in clinicalremission. One patient had switched to adalimumab treatmentbecause of IFX-related vasculitis, the second had recently under-gone an ileocecal resection, and in the third luminal activity wasseen with high levels of ATI, despite mild symptoms and normalmarkers of inflammation. During follow-up remission was reachedin 34 patients (64%) after a median of 3.3 months (IQR 1.7–8.3).Ten of those subsequently lost remission after a median of4.7 months (IQR 2.7–10.3). Survival analysis demonstrates a50% maintained remission rate at 24 months (Fig. 1). Eleven ofthe remaining 19 patients did not reach remission but did reachclinical response (Table 2).

Adalimumab Failure

During the observation period 18 patients (34%) failedadalimumab therapy, 11 of whom discontinued therapy and 7 failedbecause CD-related surgery was performed. None of the patientsdiscontinued therapy because of remission. Six patients had firstreached remission but lost response (n¼ 4) or experienced adverseeffects (n¼ 2), another 6 patients had reached clinical response butlost response (n¼ 6), and the remaining 6 patients did not respond toadalimumab and failed owing to nonresponse (n¼ 4), adverseeffects (n¼ 1), or lost response (n¼ 1)—the latter patient had milddisease activity at baseline, IFX was stopped because of an allergic

JPGN � Volume 60, Number 2, February 2015

pyright 2015 by ESPGHAN and NASPGHAN. Un

reaction, and mild activity was retained with adalimumab for12 months until a relapse occurred.

www.jpgn.org

Adalimumab treatment failed or discontinued within amedian of 5.3 months (IQR 2.9–18.0), the 4 nonresponders dis-continued within a median of 3.2 months (IQR 1.4–4.0), the 11patients who lost response failed within a median of 7.5 months(IQR 3.5–18.4), the 3 patients with adverse effects within a medianof 16 months (1.5, 16.1, and 19.6). Adalimumab failure over time isdisplayed using a cumulative hazard curve in Fig. 2: 24% failedwithin 12 months and 42% within 24 months.

Twelve patients were evaluated for antibodies-to-adalimu-mab (ATA) formation at least once during follow-up. In 4 patientsATA were present, 3 patients experienced loss of response, and thelast experienced adverse effects (fatigue after injections, hair loss,and pain and redness at the injection site).

Subanalyses

Adalimumab was less effective in patients who had notresponded to IFX than in those who had lost response to IFX. Only1 of the 3 patients with nonresponse to IFX reached remissionduring follow-up versus 24 of 34 patients who had lost response toIFX (1/3 [33%] vs 24/34 [71%], P¼ 0.24). Furthermore, 2 of theformer patients developed adalimumab failure (nonresponse andloss of response) compared with 8 of the latter patients (2/3 [67%]vs 8/34 [24%], P¼ 0.17; hazard ratio [HR] 18.8, 95% confidenceinterval [CI] 1.1–304) (Fig. 3).

ATI presence had been studied in 38 of 53 patients (72%) andfound present in 21. We detected a trend towards a higher remissionrate in patients with ATI than in those without ATI at the time ofIFX failure (17/21 [81%] vs 9/17 [53%], P¼ 0.09) and a trendtowards a lower failure rate (4/21 [19%] vs 7/17 [41%], P¼ 0.13;HR 0.37, 95% CI 0.11–1.23) (Fig. 4). Furthermore, none of theformer patients experienced nonresponse to adalimumab comparedwith 4 of the latter patients (24%). At 12 and 24 months, 14% and22% of the former patients had failed adalimumab versus 40%and 55% of the latter patients.

No differences in remission or failure rates were foundbetween patients who did and did not receive induction treatment(remission: 26/39 [67%] vs 8/14 [57%], P¼ 0.54; failure: 12/39

and 50% after 24 months.

authorized reproduction of this article is prohibited.

[31%] vs 6/14 [43%], P¼ 0.31; HR 0.5, 95% CI 0.16–1.57) orbetween patients who did and did not receive concomitant

207

Co

within 66 PY, 1.5 SAE/100 PY) was lower than that withinpreviously published studies. Within the study of Russell et al

TABLE 2. Response to adalimumab during follow-up

Follow-up, mo 0 1 4 8 12 24 36 48

Remission 3 (6) 11 (21) 18 (38) 20 (57) 16 (53) 9 (47) 4 (67) 1 (100)

Response but no remission – 20 (38) 8 (17) 6 (17) 4 (13) 4 ((21) 1 (17) –

No response – 17 (32) 7 (15) 3 (9) 2 (7) 1 (5) – –

Loss of response – – 7 (15) 6 (17) 8 (27) 5 (26) – –

Missing evaluation – 5 (9) 7 (15) – – – 1 (17) –

Total of patients on adalimumab treatment 53 53 47 35 30 19 6 1No more follow-up – 0 0 7 11 20 29 34

Adalimumab failure – 0 6 11 12 14 18 18

The results are presented as number (%¼N/total of patients on continued therapy).

100

ailu

re %

de Ridder et al JPGN � Volume 60, Number 2, February 2015

immunomodulators at baseline (remission: 22/32 [69%] vs 12/21[57%], P¼ 0.28; failure: 9/32 [28%] vs 9/21 [43%], P¼ 0.21; HR0.7, 95% CI 0.28–1.95).

AEs

In 21 patients (40%), a total 37 AEs were recorded, beingrelated to adalimumab by the treating physician, of which 14 (38%)where infections (Table 3). Only 1 serious AE (SAE) occurred, asepsis and meningitis secondary to a sinusitis (1 SAE per 66 PY, 1.5SAE/100 PY). In 3 patients adalimumab therapy was stoppedbecause of adverse effects, in 1 patient because of fatigue, hairloss, and pain and redness at the injection site, in the second becauseof discomfort after injections and recurring upper respiratory tractinfections, and vasculitis-associated skin manifestation in the third.

DISCUSSIONIn this nationwide, observational study, adalimumab therapy

induced remission in two-thirds of the IFX refractory patients, ofwhom 50% maintained remission up to 2 years. Adalimumab failureoccurred in 24% within 1 year and in 42% within 2 years. Only 1SAE occurred.

The remission rates correspond well with those in previouslypublished retrospective studies by Rosh et al (13) and Russell et al

pyright 2015 by ESPGHAN and NASPGHAN. Un

(14), but less with those in the prospective study by Hyams et al(12). The latter authors reported lower remission rates during

6 12 18 24

0

24

42

100

Ada

limum

ab fa

ilure

%

Follow-up mo

Months

Patients at risk

0

53

6

35

12

27

18

20

24

11

FIGURE 2. Adalimumab failure over time. Cumulative longitudinaladalimumab failure: within 12 months 24% of the patients failed

adalimumab therapy, rising to 42% within 24 months.

208

follow-up than the authors of the retrospective studies did. Thisdiscrepancy may be the result of a difference in denominator, asHyams et al divided the number of patients in remission by thenumber of included patients, whereas the retrospective studies,including the present study, divided it by the number of patientsstill receiving therapy. Combining the retrospective data, adalimu-mab induced remission in 61% to 64% of the IFX refractory patientswithin a median of 2.4 to 3.3 months, and after 12 months oftreatment 41% to 53% of the patients still receiving adalimumabwere in remission.

The adalimumab failure rate in our cohort corresponds wellwith that in the retrospective studies, and also with that in theprospective study. Overall, adalimumab failure occurred in 15% to28% of the IFX refractory patients within the first year and in 20%to 42% within 2 years.

Finally, the incidence of SAEs within our cohort (1 SAE

authorized reproduction of this article is prohibited.

05 12 24

19

32

Ada

limum

ab f

Follow-up mo

IFX non-responders

Patients at risk Months

IFX loss of responders

Hazard ratio 18.8 (1.1−303.6)

3

34

0

27

0

19

0

0 5 12 24

6

FIGURE 3. Comparing the risk of adalimumab failure in nonrespon-

ders versus patients who lost response to IFX. Kaplan–Meier analysis

displaying adalimumab failure over time separately for patients thathad not responded to or had lost response to infliximab. Within 4.7

months, 2 of 3 nonresponders failed adalimumab therapy; for the

remaining 1 patient no more follow-up was available. Patients with

nonresponse to IFX had higher risk for adalimumab failure thanpatients who had lost response (HR 18.8, 95% CI 1.1–303.6).

CI¼ confidence interval; HR¼hazard ratio; IFX¼ infliximab.

www.jpgn.org

Co

012 24

1424

40

55

100

Prior ATI formation

Patients at risk Months

No proper ATI formation

Hazard ratio 0.37 (0.11−1.23)

21

17

11

6

5

2

0 12 24

Ada

limum

ab fa

ilure

%

Follow-up mo

FIGURE 4. Comparing the risk of adalimumab failure between those

with versus without prior ATI formation. Kaplan-Meier analysis dis-playing adalimumab failure over time in relation to prior ATI status.

After 12 and 24 months, 14% and 22% of the patients with prior ATI

formation failed adalimumab versus 40% and 55% of the patients

without ATI formation at the time of IFX failure (HR 0.37, 95% CI

TABLE 3. Adverse events

Events

Severe infections Sinusitis/sepsis/meningitis 1

Mild infections Recurring conjunctivitis 3

Upper respiratory tract infection 3

Candida: oral 2

Candida: peristomal 1

Vasculitis 1

Otitis media 1

Herpes zoster 1

Vaginal infection 1

Other Injections painful 5

Injection site irritation 4

Eczema or other skin disorders 3

Discomfort after injection 2

Hair loss 2

Arthralgia 2

Tiredness 2

Headache 1

Dizziness 1

Myalgia 1

Total number of AEs 37

No. inflicted patients 21 (40%)

In total, 37 adverse events were reported in 21 patients (40%), of which38% where infections. Only 1 severe adverse event was reported, a severe

JPGN � Volume 60, Number 2, February 2015 Adalimumab Therapy in Children With IFX-Refractory CD

4 SAEs were reported within 72.5 PY (5.5 SAE/100 PY) and in theprospective study 63 SAEs developed within 152 PY (41 SAE/100PY). The higher incidence in the prospective study may be the resultof the use of more extensive criteria for SAE and morestringent monitoring.

Efficacy in IFX Nonresponders Versus Loss ofResponders

Within our cohort, we have demonstrated that a small groupof patients who had not responded to IFX (n¼ 3), had a higher riskfor adalimumab failure than those who had lost response to IFX.Some patients do not respond well to anti-TNF therapy, whereasother patients do, which may be caused by a difference in under-lying disease mechanism. Different disease mechanisms mayrequire different treatment strategies.

This issue has not been studied previously in paediatricpatients with CD, only in adult patients with IBD. Ho et al (16)studied patients with CD and found a trend for lower remission ratesand higher failure rates in IFX nonresponders compared with initialresponders (percentages and significance not clarified). Garcia-Bosch et al (17) studied patients with ulcerative colitis and founda difference in response rates after 12 weeks of treatment betweenIFX prior nonresponders (2/6, 33%) and initial responders (25/33,76%, P¼ 0.01). Both studies affirm the difference in efficacy foundin our study.

Efficacy in Patients With Versus Without ATI

We found a trend for higher remission rates and lower failurerates in patients with ATI than in those without. A possibleexplanation for this efficacy difference is this: patients initiallyresponding to the first anti-TNF agent, who are confronted with loss

0.11–1.23). ATI¼ antibodies to IFX; CI¼ confidence interval; HR¼hazard ratio; IFX¼ infliximab.

pyright 2015 by ESPGHAN and NASPGHAN. Un

of response resulting from antidrug antibodies, evade the causativefactor for loss of response and regain response by switching to

www.jpgn.org

another anti-TNF agent. On the contrary, patients without antidrugantibodies at the time of anti-TNF treatment failure fail for anotheryet unknown reason, which makes regaining response more chal-lenging. Failure owing to ATI therefore seems more favourablebecause a response can be more easily regained.

This issue has not been studied previously in paediatricpatients with CD, only in adult patients with IBD. In the GAINtrial (Gauging Adalimumab Efficacy in Infliximab Nonresponders)no difference was seen in remission rates after 4 weeks of adali-mumab therapy between patients with CD with and without priorATI before adalimumab (11/50 [22%] vs 19/88 [22%]) (11). Westet al (18) reported no relation between the presence of ATI andresponse to adalimumab in patients with CD, but serum ATI levelswere higher in adalimumab nonresponders than in responders,suggesting a negative influence of high ATI concentrations onadalimumab response. Afif et al (19) found a trend for increasedresponse in ATI-positive patients with ulcerative colitis versusATI-negative patients (5/8 (63%) vs 1/4 (20%), P� 0.5).

Strengths and Limitations

This study is the first nationwide study in paediatric patientswith CD evaluating adalimumab therapy in clinical practice and isstrengthened by its population-based design. Because of its obser-vational character the study reflects daily practice and its nation-wide approach provides the inclusion of a broad spectrum ofpatients. The retrospective design, however, has several limitations.

First, the presented data may not be complete. Althoughparticipating centres were asked to identify all of the eligiblepatients, some may have been missed. When selectively more orless severely diseased patients were missed, the reported efficacymay differ from the efficacy in the total population. Some non-

infection. AE¼ adverse event.

authorized reproduction of this article is prohibited.

SAEs may also have been missed, such as minor infections treatedin primary care practice.

209

Co

Second, the efficacy assessment was suboptimal. We usedclinical disease activity indices and not mucosal healing or markersfor mucosal healing to evaluate adalimumab efficacy. Furthermore,the wPCDAI could not be calculated at each time point and a PGAwas used in the remainder.

Finally, not all of the patients were treated in the same way.Patients received different adalimumab induction dosages anddifferent concurrent medication. Decisions concerning their treat-ment were based on the judgement of the treating physician, as weredecisions concerning treatment cessation. Variations in confoun-ders may have biased our outcome data and the subanalyses.

CONCLUSIONSThis study demonstrates that in clinical practice adalimumab

is an effective therapy for IFX refractory paediatric patients withCD with only limited adverse effects. We therefore recommend itsusage in these otherwise difficult-to-treat patients. Well-powered,long-term pharmacovigilance studies are needed to further establishthe safety of adalimumab therapy in paediatric patients with CD,especially regarding late-onset AEs, such as malignancies.

Adalimumab appears to be less effective in the treatment ofIFX nonresponders and patients previously failing IFX without thepresence of ATI. To further establish these efficacy differences,they should be confirmed in larger cohorts. Because of its clinicalrelevance, research should attempt to elucidate possible differencesin disease mechanism, so that disease mechanism specific therapycan be given, and costly, ineffective therapy can be avoided.

REFERENCES1. Kelsen J, Baldassano RN. Inflammatory bowel disease: the difference

between children and adults. Inflamm Bowel Dis 2008;14:S9–11.

2. van der Zaag-Loonen HJ, Casparie M, Taminiau JA, et al. The incidenceof pediatric inflammatory bowel disease in the Netherlands: 1999–2001. J Pediatr Gastroenterol Nutr 2004;38:302–7.

3. Vasseur F, Gower-Rousseau C, Vernier-Massouille G, et al. Nutritionalstatus and growth in pediatric Crohn’s disease: a population-based study.Am J Gastroenterol 2010;105:1893–900.

4. Van Assche G, Dignass A, Reinisch W, et al. The second Europeanevidence-based Consensus on the diagnosis and management of Crohn’sdisease: Special situations. J Crohns Colitis 2010;4:63–101.

5. Hyams J, Crandall W, Kugathasan S, et al. Induction and maintenance

de Ridder et al

pyright 2015 by ESPGHAN and NASPGHAN. Un

disease in children. Gastroenterology 2007;132:863–73.

210

6. De Bie CI, Hummel TZ, Kindermann A, et al. The duration of effect ofinfliximab maintenance treatment in paediatric Crohn’s disease islimited. Aliment Pharmacol Ther 2011;33:243–50.

7. Hyams JS, Lerer T, Griffiths A, et al. Long-term outcome of main-tenance infliximab therapy in children with Crohn’s disease. InflammBowel Dis 2009;15:816–22.

8. Rutgeerts P, Van Assche G, Vermeire S. Review article: Infliximabtherapy for inflammatory bowel disease—seven years on. AlimentPharmacol Ther 2006;23:451–63.

9. Cassinotti A, Travis S. Incidence and clinical significance of immuno-genicity to infliximab in Crohn’s disease: a critical systematic review.Inflamm Bowel Dis 2009;15:1264–75.

10. Colombel JF, Sandborn WJ, Rutgeerts P, et al. Adalimumab for main-tenance of clinical response and remission in patients with Crohn’sdisease: the CHARM trial. Gastroenterology 2007;132:52–65.

11. Sandborn WJ, Rutgeerts P, Enns R, et al. Adalimumab induction therapyfor Crohn disease previously treated with infliximab: a randomised trial.Ann Intern Med 2007;146:829–38.

12. Hyams JS, Griffiths A, Markowitz J, et al. Safety and efficacy ofadalimumab for moderate to severe Crohn’s disease in children. Gastro-enterology 2012;143:365.e2–74.e2.

13. Rosh JR, Lerer T, Markowitz J, et al. Retrospective Evaluation of theSafety and Effect of Adalimumab Therapy (RESEAT) in pediatricCrohn’s disease. Am J Gastroenterol 2009;104:3042–9.

14. Russell RK, Wilson ML, Loganathan S, et al. A British Society ofPaediatric Gastroenterology, Hepatology and Nutrition survey of theeffectiveness and safety of adalimumab in children with inflammatorybowel disease. Aliment Pharmacol Ther 2011;33:946–53.

15. Turner D, Griffiths AM, Walters TD, et al. Mathematical weighting ofthe pediatric Crohn’s disease activity index (PCDAI) and comparisonwith its other short versions. Inflamm Bowel Dis 2012;18:55–62.

16. Ho GT, Mowat A, Potts L, et al. Efficacy and complications ofadalimumab treatment for medically-refractory Crohn’s disease: ana-lysis of nationwide experience in Scotland (2004–2008). AlimentPharmacol Ther 2009;29:527–34.

17. Garcia-Bosch O, Gisbert JP, Canas-Ventura A, et al. Observational studyon the efficacy of adalimumab for the treatment of ulcerative colitis andpredictors of outcome. J Crohns Colitis 2013;7:717–22.

18. West RL, Zelinkova Z, Wolbink GJ, et al. Immunogenicity negativelyinfluences the outcome of adalimumab treatment in Crohn’s disease.Aliment Pharmacol Ther 2008;28:1122–6.

19. Afif W, Leighton JA, Hanauer SB, et al. Open-label study of adalimu-mab in patients with ulcerative colitis including those with prior loss ofresponse or intolerance to infliximab. Inflamm Bowel Dis 2009;15:1302–7.

20. Levine A, Griffiths A, Markowitz J, et al. Pediatric modification of the

JPGN � Volume 60, Number 2, February 2015

Montreal classification for inflammatory bowel disease: the Paris clas-

infliximab therapy for the treatment of moderate-to-severe Crohn’s

authorized reproduction of this article is prohibited.

sification. Inflamm Bowel Dis 2011;17:1314–21.

www.jpgn.org