Profile of pediatric Crohn's disease in Belgium

Ava i l ab l e on l i ne a t www.sc i enced i r ec t . com

Journal of Crohn's and Colitis (2013) xx, xxx–xxx

CROHNS-00779; No of Pages 11

Profile of pediatric Crohn's disease in Belgium☆,☆☆

E. De Greef a, b,⁎, J.M. Mahachie John c, d, I. Hoffman e, F. Smets f,S. Van Biervliet g, M. Scaillon h, B. Hauser b, I. Paquot i, P. Alliet j, W. Arts k,O. Dewit l, H. Peeters m, F. Baert n, G. D'Haens o, J.F. Rahier p, I. Etienne q,O. Bauraind r, A. Van Gossum s, S. Vermeire t, F. Fontaine u, V. Muls v,E. Louis w, F. Van de Mierop x, J.C. Coche y, K. Van Steen c, d,G. Veereman a, bfor the IBD working group of the Belgian Society of PediatricGastroenterology, Hepatology and Nutrition (BeSPGHAN) and the Belgian IBDResearch and Development (BIRD)

a Pediatric Gastroenterology, Queen Paola Children's Hospital, Antwerp, Belgiumb Pediatric Gastroenterology, UZB, Brussels, Belgiumc Systems and Modeling Unit, Montefiore Institute, ULG, Liege, Belgiumd Bioinformatics and Modeling, GIGA-R, ULG, Liege, Belgiume Pediatric Gastroenterology, UZ Gasthuisberg, Leuven, Belgiumf Pediatric Gastroenterology, Université Catholique de Louvain, Cliniques Universitaires St. Luc, Brussels, Belgiumg Pediatric Gastroenterology, UZ Gent, Belgiumh Pediatric Gastroenterology, University Children's Hospital Queen Fabiola, Brussels, Belgiumi Pediatric Gastroenterology, CHC Clinique de l'espérance, Liège, Belgiumj Pediatric Gastroenterology, Jessa Hospital, Hasselt, Belgiumk Pediatric Gastroenterology, ZOL Genk, Genk, Belgiuml Gastroenterology, UCL St Luc, Brussels, Belgium

Abbreviations: BIRD, Belgian IBD Research and Development Group; BELCRO, Belgian Registry for Pediatric Crohn's Disease; BESPGHAN, BelgianSociety for Pediatric Gastroenterology, Hepatology and Nutrition; CRF, clinical report file; CRP, C reactive protein; CD, Crohn's disease; IBD,inflammatory bowel disease; GI, gastrointestinal; m, months; PCDAI, Pediatric Crohn's Disease Activity Index; PGA, Physician's Global Assessment; w,weeks; y, year; 5-ASA, 5-aminosalicylic acid; 6-MP, 6-mercaptopurin

☆ Conference presentations: The data described in thismanuscriptwere presented in part at the BelgianWeek of Gastroenterology, February 2011and February 2012, at the Belgian Society of Pediatrics, March 2011 and March 2012. They were presented as a poster at the meeting of the EuropeanCrohn's and Colitis Organisation in Dublin, March 2011 and in Barcelona, February 2012; at the Digestive DiseaseWeek in Chicago, May 2011 and in SanDiego May 2012. Theywere published in abstract form in the Acta Gastroenterologica Belgica, Belgisch Tijdschrift voor Kindergeneeskunde, Journal ofCrohn's and Colitis and Gastroenterology in 2011 and 2012.☆☆ Support source: This research was supported by a grant from MSD Medical, Belgium.

⁎ Corresponding author at: Pediatric Gastroenterology, Hepatology and Nutrition, UZ Brussels, Laarbeeklaan 101, 1090 Brussels, Belgium.Tel.: +32 24749145.

E-mail addresses: [email protected] (E. De Greef), [email protected] (J.M. Mahachie John), [email protected](I. Hoffman), [email protected] (F. Smets), [email protected] (S. Van Biervliet), [email protected](M. Scaillon), [email protected] (B. Hauser), [email protected] (I. Paquot), [email protected] (P. Alliet), [email protected] (W. Arts),[email protected] (O. Dewit), [email protected] (H. Peeters), [email protected] (F. Baert), [email protected] (G. D'Haens),[email protected] (J.F. Rahier), [email protected] (I. Etienne), [email protected] (O. Bauraind), [email protected](A. Van Gossum), [email protected] (S. Vermeire), [email protected] (F. Fontaine), [email protected] (V. Muls),[email protected] (E. Louis), [email protected] (F. Van de Mierop), [email protected] (J.C. Coche),[email protected] (K. Van Steen), [email protected] (G. Veereman).

1873-9946/$ - see front matter © 2013 European Crohn's and Colitis Organisation. Published by Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.crohns.2013.04.016

Please cite this article as: De Greef E, et al, Profile of pediatric Crohn's disease in Belgium, J Crohns Colitis (2013), http://dx.doi.org/10.1016/j.crohns.2013.04.016

mailto:[email protected]


























http://dx.doi.org/10.1016/j.crohns.2013.04.016

http://dx.doi.org/



2 E. De Greef et al.

m Gastroenterology, UZ Gent, Belgiumn H. Hart Hospital, Roeselare, Belgiumo Gastroenterology, Imelda Hospital, Bonheiden, Belgiump UCL Mont Godinne, Mont Godinne, Belgiumq Pediatric Gastroenterology, CHR de la Citadelle, Liège, Belgiumr Pediatric Gastroenterology, Clinique St Pierre, Ottignies, Belgiums Gastroenterology, ULB Erasme Hospital, Brussels, Belgiumt Gastroenterology, UZ Gasthuisberg, Leuven, Belgiumu Gastroenterology, CHU Saint Joseph, Liège, Belgiumv Gastroenterology, CHU St Pierre, Brussels, Belgiumw Gastroenterology, CHU and University of Liège, Belgiumx Gastroenterology, St. Augustinus Hospital, Antwerp, Belgiumy Gastroenterology, Clinique St Pierre, Ottignies, Belgium

Received 24 August 2012; received in revised form 12 April 2013; accepted 13 April 2013

Please cite this article as: De Greef E,j.crohns.2013.04.016

KEYWORDSPediatric;Crohn's disease;Registry;Diagnosis;Profile;Children;Disease phenotype

Abstract

Aim: A Belgian registry for pediatric Crohn's disease, BELCRO, was created. This first report aims atdescribing disease presentation and phenotype and determining associations between variables atdiagnosis and registration in the database.Methods: Through a collaborative network, children with previously established Crohn's diseaseand newly diagnosed children and adolescents (under 18 y of age) were recruited over a 2 yearperiod. Data were collected by 23 centers and entered in a database. Statistical association testsanalyzed relationships between variables of interest at diagnosis.
Results: Two hundred fifty-five patients were included. Median age at diagnosis was 12.5 y (range:1.6–18 y); median duration of symptoms prior to diagnosis was 3 m (range: 1–12 m). Neonatal historyand previous medical history did not influence disease onset nor disease behavior. Fifty three % ofthese patients presented with a BMI z-score b −1. CRP was an independent predictor of diseaseseverity. Steroids were widely used as initial treatment in moderate to severe and extensive disease.Over time, immunomodulators and biological were prescribed more frequently, reflecting a lowerprescription rate for steroids and 5-ASA. A positive family history was the sole significant determinantfor earlier use of immunosuppression.Conclusion: In Belgium, the median age of children presenting with Crohn's disease is 12.5 y.Faltering growth, extensive disease and upper GI involvement are frequent. CRP is an independentpredictive factor of disease activity. A positive family history appears to be the main determinant forinitial treatment choice.© 2013 European Crohn's and Colitis Organisation. Published by Elsevier B.V. All rights reserved.
et al, Profile of pediatric Crohn's

1. Introduction

The incidence of Crohn's disease (CD) increases, especially inwesternized countries.1 Approximately 25% of patients presentduring childhood.2 In children a more severe and extensivedisease phenotype is described compared to adults.3 Theimpact on the child's growth and development is an importantfactor determining treatment strategies.

The natural course of CD remains unpredictable. Based onadult literature, risk factors for severe disease are younger ageat diagnosis, the presence of perianal disease and smoking.4 Inpediatrics, these risk factors need confirmation and otherfactors, possibly related to growth and development need to beidentified. High concordance of CD in monozygotic twins and apositive family history for inflammatory bowel disease (IBD) in5–20% confirm an underlying genetic susceptibility.5 Environ-mental influence is proven by the deleterious effect of smokingand the rise in CD in immigrant populations from regions with

dis

low prevalence to regions with high prevalence.1 Regionalinformation, captured in registries, aims at providing insights indisease presentation, disease course and influencing environ-mental factors.6,7 We therefore initiated a registry of Belgianpediatric CD patients (BELCRO). In this manuscript we report onpatient characteristics at diagnosis and for previously diagnosedpatients at inclusion in the database.

2. Materials and Methods

2.1. Population

BELCRO was initiated in May 2008 through a collaboration ofthe IBD working group of the Belgian Society for PediatricGastroenterology, Hepatology and Nutrition (BESPGHAN) andthe Belgian IBD Research and Development Group (BIRD). Theaim of the registry is to describe a cohort of old and newly

ease in Belgium, J Crohns Colitis (2013), http://dx.doi.org/10.1016/



3

diagnosed pediatric CD patients recruited over a 2 y periodand to prospectively follow them for 5 y. All Belgian pediatricand adult gastroenterology centers were invited to participatein the registry. All the CD patients under their care and all thenewly diagnosed CD patients were asked to participate in theregistry. Twenty-three pediatric and adult units, representingall major Belgian centers and members of the scientificcommittees have recruited their patients (6 pediatric tertiarycare GI services, 6 adult tertiary care GI services, 5 pediatricand 6 adult peripheral GI services). The diagnosis of CD had tobe established according to the Porto criteria.8 Informedconsent was obtained from the parents or legal guardians,the patients gave their assent. The study protocol wasestablished following the declaration of Helsinki and GoodClinical Practice guidelines, approved by the ethics committeeZNA Middelheim, Antwerp Belgium (nr 3147) and registered onclinical trials.gov (B00920083829).

2.2. Data Collection

A chart review was performed for previously diagnosed CDpatients (diagnosed before May 1st 2008). Data at diagnosisand at inclusion in the registry were extracted from themedical files into a standardized clinical record file (CRF).Data from newly diagnosed patients (diagnosed after May 1th2008) were immediately collected in the CRF. All data wereentered into an Excell® database (Microsoft Corporation,Washington, USA) by a dedicated data manager (DRC Datamanagement, Gent, Belgium). The procedure for data collec-tion and ownership was described in a Charter by a steeringcommittee comprising the principal investigator and represen-tatives of the participating scientific societies.

2.3. Description of Variables

The following information was collected from all patients atdiagnosis: demographics (race, age, gender), neonatal history(mode of delivery, birth weight, gestational age, mode offeeding), family history (CD, ulcerative colitis, auto-immunediseases), previous medical history (infections, surgery,stressful events, food allergies) and concomitant conditions(hepatitis, celiac disease, psoriasis, lupus), symptoms andsigns at presentation (abdominal pain, diarrhea, perianaldisease, extra-intestinal manifestations), diagnostic work-up(including laboratory, endoscopy, histology and imaging) andtreatment. Upon inclusion in the database, data on symptoms,vaccinations, therapy, concomitant conditions and laboratoryvalues were recorded from previously diagnosed patients.Age categories (b6 y, ≥6 y –≤ 12 y, N12 y) were defined tofurther stratify the population.

Disease severity was scored using the Pediatric Crohn'sDisease Activity Index (PCDAI), a validated scoring systembased on symptoms, biochemical parameters, clinical examand growth.8 When PCDAI was not available, Physician'sGlobal Assessment (PGA) was obtained, based on the clinicalevaluation of the patient by his physician. Faltering growthwas considered a height = 5 or 10 within the PCDAI. Diagnosticprocedures were recorded and the involved intestinal areas ordisease location were derived from endoscopic data, histologyand/or imaging. Disease location was classified following theMontreal classification (ileal (L1), colonic (L2), ileocolonic

Please cite this article as: De Greef E, et al, Profile of pediatric Crohn's disj.crohns.2013.04.016

(L3), upper gastrointestinal (GI) (L4))9 as well as by the morerecently published Paris classification10 (L4A upper GI involve-ment until the angle of Treitz and L4B upper GI involvementbeyond the angle of Treitz). Data on initial treatment werecollected and stratified in the following categories: enteralnutrition, 5 ASA, antibiotics, steroids (budesonide, predniso-lone), immunomodulators (6 mercapotpurine, methotrexate,azathioprine), biologicals (infliximab, adalumimab), tacroli-mus and cyclosporine.

2.4. Statistical Analysis

All data were arranged and processed for handling usingMicrosoft™ Office Excel and analyzed with SPSS 17.0. Descrip-tive statistics were used to describe the populationfeatures. Non-parametric association tests were used toinvestigate relationships between variables of interest. Inparticular, Fisher's exact tests were used to assess relation-ships between categorical variables. For continuous variableoutcomes and categorical explanatory variables, Mann–Whit-ney U tests (Kruskal–Wallis tests when N2 categories) wereperformed. In addition, multiple regression analyses (linear andlogistic analyses) were carried out whenever appropriate. Alltests were performed on available cases only. Tests werecarried out at a significance level of 5%. To avoid bias,previously diagnosed patients (beforeMay 1th 2008) and newlydiagnosed patients (after May 1th 2008) were compared fortheir differences and a separate analysis was performed forthe differing factors. Logistic regression models were fitted toinvestigate the association between the 2 groups of patientsand the recorded variables. Nomultiple testing correction wasperformed. Non-uniform variables were used for all analyses,making it difficult to generalize conclusions across outcomes.

3. Results

3.1. Demographics and Neonatal Data

Two-hundred fifty-five patients under 18 y at diagnosis (bornafter May 1st 1990), with established diagnosis of CD accordingto the Porto criteria11 and living in Belgium,were recruited overa 2 y period (May 1st 2008–April 30th 2010). In this population100/255 patients were newly diagnosed and 155/255 werepreviously diagnosed. Male/female ratio was 1.2 and 98% ofpediatric patients were Caucasian. The remaining 2% was ofAsian or South American origin. Neonatal data reported amedian birth weight of 3.3 kg (range 1.4–4.6 kg), a mediangestational age of 40 w (range: 28–42 w). Caesarian sectionoccurred in 29 patients (11%) and 78% were exclusively orpartially breast fed for a median duration of 7 w (range 0–140 w). In comparing newly diagnosed and previously diagnosedpatients no significant difference appeared for the demographicand neonatal data.

3.2. Previous Medical History

Seventy percent of patients were diagnosed by a pediatricgastroenterologist and just more than half in a universitycenter. History taking revealed that 23% of patients hadsuffered from a recent infectious episode. In 9.2% there was




0 20 40 60 80 100

Diarrhoea

Abdominal Pain

Weight Loss

Growth Retardation

Extra Intestinal Manifestations

Perianal Disease

%Patients

Figure 1 Presenting symptoms at diagnosis.


a recent history of an episode of self limiting diarrhea that hadbeen labeled as acute gastro-enteritis. There were no reportsof proven bacterial enteritis prior to diagnosis. Twenty-three% experienced a major stressful event (i.e. defined as beingstressful for the patient by the patient himself, his parents andthe physician) and 42% had past surgery, of which ear–nose–throat surgery in 45% and inguinal hernia correction andcircumcision in 15%. Eight patients had previous appendecto-my (7%). Disease related surgery at diagnosis was reported in17%, consisting of abscess drainages and surgical fistulatreatment. In previously and newly diagnosed patients 24%and 7% respectively had disease related surgery prior todiagnosis. With this exception, no significant differences werefound in the medical history of the previously and newlydiagnosed patients.

Data on immunization were available from 248 patients,although not all data was complete. Polio, the only obligatoryvaccination in Belgium, was recorded in 222 patients (86%).Fifty three patients received vaccines in addition to thenational recommendations,12,13 including hepatitis A, yellowfever, influenza and typhoid. Six patients were vaccinated forvaricella.

Passive smoking was present in 16% of patients, activesmoking in 1%.

3.3. Family History

A positive family history for IBD in first degree relativesoccurred in 29 patients (11.4%). CD was mentioned in 25patients, ulcerative colitis in 4. Both conditions weremutuallyexclusive. Auto-immune pathology, including psoriasis, lupus,diabetes type I and celiac disease affected the broader family(parents, siblings, grandparents, cousins, aunts and uncles) of35.9% of patients. Family history was similar in previously andnewly diagnosed patients.

3.4. Presentation

Belgian pediatric CD patients presented at a median age of12.5 y (range 1.6–18.0 y) after a median duration of symptomsof 3 m (range 1–12 m). Median symptom duration prior todiagnosis did not differ for patients with a positive familyhistory. Symptoms at diagnosis are presented in Fig. 1. Themainpresenting symptoms were abdominal pain (84%), diarrhea(72%) and weight loss or lack of weight gain (72%).

Height z-scores and BMI z-scores were available from240 patients. Severe growth retardation at diagnosis(z-score b −2SD) affected 8.7% of the population at diagnosis.Sixty patients (25%) had a BMI z-score ≤ −2SD. The previouslydiagnosed group belonged mostly to the age category N12 ywhereas median and mean age were comparable for previousand recent diagnoses. The previously diagnosed patients hadlower z-scores for height compared to newly diagnosedpatients. (−0.52 (range −5.35 to 5.87) vs. −0.23 (range −3.40to 12.71); p = 0.016).

3.5. Disease Location and Severity at Diagnosis

Disease location was classified according to the Montreal andParis classification. Results are shown in Table 1. Diagnosis ofIBD had to fulfill the Porto criteria, however not all participating


centers performed routine upper endoscopy and small bowelimaging. Hundred and ninety one of 255 patients underwentupper endoscopy and the ileum was not visualized in 12patients. Small bowel involvement was not evaluated in allpatients. In those who had a small bowel evaluation differenttechniques were used: 38 by CT abdomen, 33 by MRI, 65 bycontrast studies, 10 by white blood cell scan and 1 by capsuleendoscopy. In total, the upper GI tract was evaluated byimaging and/or endoscopy in 205/255 patients and wasinvolved in 70%. Isolated ileal disease (L1) was present in 32(13%), isolated colonic involvement (L2) in 62 (24%) patients,ileocolonic disease (L3) in 157 (61%) and isolated upper GIinvolvement in 4 patients (2%).

The Paris classification subdivides upper GI involvement inL4A (proximal to Treitz ligament) and L4B (distal to Treitzligament). These regions were involved in 96/144 and 35/144patients respectively. Forty eight patients had both L4A and L4Binvolvement.

Perianal disease at diagnosis, as described in the PCDAI,occurred in 28% of patients. Stricturing disease wasmentioned in 15 patients (5.8%) at diagnosis. Out of the245 patients in whom extraintestinal manifestations wererecorded (erythema nodosum, arthritis, uveitis, arthral-gias, pyoderma gangrenosum), 72 scored positive (29%). FullPCDAI was obtained for 134/255 patients. Physicians takingcare of adult patients are not used to PCDAI. Whenunavailable, the physician was asked for a PGA for eachpatient. The results are presented in Table 2. Disease wasmild in 24%, moderate in 43% and severe in 28% of thecohort. Disease severity and disease location were compa-rable in previously and newly diagnosed patients.

3.6. Therapy

Treatment at diagnosis was classified according to thefollowing categories: enteral therapy, 5-ASA, antibiotics,systemic steroids (prednisolone, budesonide), immuno-modulators (6MP, methotrexate, azathioprine), biologi-cals (infliximab, adalumimab), cyclosporine and tacrolimus(Table 3). At diagnosis, monotherapy was initiated in 23.9% ofpatients, mainly steroids (10.9%) or 5 ASA (9%). In the majorityof patients, a combination of several drugs was initiated. Oralsteroids, immunomodulators and 5 ASA were part of the initialcombination treatment in respectively 64%, 43% and 40%. In thepreviously diagnosed patients, the proportion of patients on5-ASA as initial treatment was higher compared to the newlydiagnosed patients.




Table 1 Disease location at diagnosis and per age group following the Montreal/Paris classification.

Montreal Paris

Disease location L1 L2 L3 Isolated L4 L4 L4A L4BTotal cohort(n = 255)

32(12.5%)

62(24%)

157(61.5%)

4(2%)

179(70%)

144(56%)

83(32%)

Not evaluated 12 0 0 5 5 6 20b6 y (n = 10) a 3 (30%) 4 (40%) 3 (30%) a 0 8 (80%) 8 (80%) 2 (20%)*6–11 y (n = 91) a 7 (8%) 24 (26%) 59 (65%) a 1(1%) 67 (74%) 56 (62%) 31 (37%)*12–18 y (n = 153) a 22 (14.4%) 34 (22.2%) 94 (61.4%) a 3 (2%) 104 (68%) 80 (53%) 49 (34%)*a In 1 patient, exact age at diagnosis was missing.

5

3.7. Previously Diagnosed Patients at Time of Inclusionin the Database

For the 155/255 previously diagnosed patients the mediandisease duration at registration in the databasewas 2.7 y (range0.3–8.2 y) with a median age at registration of 15.9 y (range:5.3–19.8 y). The majority of patients had inactive disease(70.4%) and was asymptomatic. Abdominal pain was men-tioned in 28.2%, diarrhea in 12.4%, weight loss in 12.5%,perianal disease in 7.2% and extra-intestinal manifestationsin 10.5%. Twenty patients had undergone surgery during theirinitial disease course of which 17 were disease related: 3abscesses, 3 fistulectomies, 6 ileo-caecal resections and 1 smallbowel resection, 3 colonic resections and 1 fissure treatment.The majority (84.2%) had received combination treatmentincluding steroids (65.1%), immunomodulators (65.8%), 5-ASA(48.7%) and/or biologicals (28.9%). At the date of inclusionin the registry only 18/155 patients were still on steroids (9budesonide, 9 prednisone), while their disease activity wasinactive in 8/18, mild in 7/18 and moderate in 3/18. In 14 otherpatients start and stop dates of steroid therapy were imprecise,so they possibly had ongoing treatment.

3.8. Associations at Diagnosis (Table 4)

The analysis was carried out on the entire group for variablesthat did not differ between previously and newly diagnosedpatients. A separate analysis was performed for the differingvariables in order to avoid bias due to group heterogeneity.

Neonatal variables and medical history revealed no associ-ations with age, disease location or disease severity. Breastfedpatients were often diagnosed at a younger age (p = 0.0003)and tended to have more colonic disease (L2) (p = 0.014).Patients diagnosed at a younger age more often mentioned an

Table 2 Severity of disease at diagnosis per age group, scored b

Disease severity Inactive Mild

Total cohort (n = 255) 10(4%)

61(24%)

b6 y (n = 10) 0(0%)

5(50%)

6–11 y (n = 90) 2(2.2%)

24(26.7%)

12–18 y (n = 152) 8(5.3%)

32(21%)


affected family member (p = 0.032). Age was not related withdisease location for the whole group, but in the prospectivecohort, upper GI involvement (L4A) and in the previouslydiagnosed patients L3 involvement was seen in older patients(p = 0.047; p = 0.057). Even though univariate analysis showedan association between disease location (L1, L3) and heightz-scores with disease severity, multiple regression analysiswithheld only CRP as independent predictive factor for diseaseseverity.

In the total cohort, several disease related factors wereassociated with the initial treatment choice. Because ofthe limited number of patients on monotherapy at diagno-sis, associations found in this group have little relevance.Young age at diagnosis was associated with a more frequentuse of antibiotics (p = 0.002), steroids and 5 ASA (p = 0.001;p = 0.004) as part of their initial treatment. Steroids incombination treatment and enteral supplements were linkedwith moderate to severe disease or L4 involvement (p = 0.045and p = 0.048);(p = 0.005 and p = 0.043), where in patientswith L1 involvement antibiotics were more often associated(p = 0.032). In the prospective cohort, steroids were less likelyused in patients with perianal disease (p = 0.006) and patientswith lower z-scores for height weremore often started enteralsupplements at diagnosis (p = 0.002).

3.9. Associations Between Variables at Diagnosis andat Inclusion for Previously Diagnosed Patients (Table 5)

In the previously diagnosed group, follow-up data obtained atregistration in the database were also analyzed. We noticedan important effect of the disease duration on several factors.Patients with a longer disease course at inclusion werediagnosed at a younger age (b0.001) and had better z-scoresfor height at diagnosis (p = 0.008). They were more likely to

y PCDAI or PGA.

Moderate Severe missing

110(43%)

71(28%)

3(1%)

3(30%)

2(20%)

39(43.3%)

25(27.8%)

68(44.7%)

44(29%)




Table 3 Medication at diagnosis (n = 255).

Monotherapy Combination therapy

5-ASA 9% 40.6%Enteral nutrition 0.4% 8.2%Rectal therapy 0.4% 4.3%Immunomodulators 1.6% 43%Biologicals 0% 1%Steroids 10.9% 64.8%AB 1.6% 25%Tacrolimus/cyclosporin 0% 1%


have had 5 ASA at diagnosis (p b 0.001). An association wasfound between the z-scores for height at diagnosis and BMIz-scores at inclusion (p = 0.025) with weight as interferingfactor indicating a better weight gain over time compared togrowth.

Patients who received steroids at diagnosis were morelikely to have a better disease control at the time of inclusion(p = 0.004; OR = 3.8 (95%CI 1.7–8.4)). Patients with L1received less steroids at diagnosis and had less diseasecontrol at inclusion (p = 0.03; OR 0.3 — 95%CI 0.1–0.9)compared to patients with L3 who had a better diseasecontrol at inclusion (p = 0.02; OR 2.8 — 95%CI 1.1–7.4).Need for surgery was only influenced by disease behavior (S)(p = 0.001; OR 6.8 — 95% CI 1.8–25.3) not by disease severity,location or other treatment modalities. No further significantassociations were found between disease related elements suchas disease severity or disease location and initial treatment,only a positive family history for IBDwas associatedwith the useof 5-ASA at diagnosis (OR 2.1 95% CI 1.07–4.49), and 5-ASA andimmunomodulators during follow up (p = 0.02; OR 2.5 — 95%CI1.1–4.3 and p = 0.004; OR 2.7 — 95%CI 1.3–5.5).

There was a decrease in the use of steroids and 5-ASA(p = 0.001; OR 0.02 — 95%CI 0.002–0.3; p = 0.001; OR 0.03— 0.004–0.3) and an increase in prescriptions of immuno-modulators (p = 0.03; OR 1.8–95%CI 0.08–41.1) over time,paralleling a decrease in disease severity (p = 0.01) (Fig. 2).The decrease in disease severity can be reflected by thedecrease in perianal disease and extra-intestinal manifes-tations between diagnosis and inclusion in the database,respectively 28% vs. 7.2% and 29% vs. 10.5%. The associ-ation between immunomodulator monotherapy at diagno-sis and a better height z-score at inclusion is based on toofew patients (4 patients).

4. Discussion

This is the first report on Belgian pediatric CD patients.Nationwide recruitment by pediatric and adult gastroenterolo-gists, members of national scientific societies BESPGHAN andBIRD, intended to reach as many pediatric patients as possibleand represents the actual care takers. Since adolescentswith IBD are often cared for by adult gastroenterologists,the collaboration between both pediatric and adult centersguarantees a better representation.

All centers included their current patients. Therefore,this group is the best possible representation of pediatricIBD in the country, as there is no national registration. Eventhough virtually all pediatric GI centers caring for pediatric


IBD patients participated in the study, it is impossible toevaluate the exact number of pediatric patients treated byadult gastroenterologists. It is however improbable thatadult gastroenterologist would treat patients below the ageof 16 y.

Obviously, data at diagnosis are more detailed in theprospective cohort. However, the retrospective group isvaluable for comparison because, among other factors, thetherapeutic strategy has evolved over time. Outcomes ofboth groups will be compared in follow up.

Since the same centers included their current patientsand retrospective cohort there is no bias. Therefore, there isno reason to assume that the previously diagnosed patientsare a selected group.

The median age of disease onset in BELCRO patients iscomparable to what is found in surrounding countries.3,7,14–20 Incontrast to the adult population, the majority of pediatricpatients are male. Female preponderance started to show fromage 15 y on. Median duration of symptoms prior to diagnosis was3 m,which is shorter than in surrounding countries.16 A possibleexplanation is that specialized medical care is easily accessi-ble in Belgium, professional referral is not mandatory, traveldistances are short and waiting lists are usually short ornon-existent. Health insurance is offered to all citizens. Thereare no public campaigns, thus general awareness of CD isprobably comparable to surrounding countries. The majorityof patients were Caucasian, including immigrants from NorthAfrica and Turkey for whom no subset was made.

Neonatal history and previous medical history did notinfluence disease severity nor disease location in this cohort.Younger age at diagnosis was associated with breastfeeding andan infectious episode prior to diagnosis. Seventy eight percentof BELCRO patients have been (partially/exclusively) breastfedin the neonatal period. These results compare to 71% ofmothersstarting breastfeeding in the general population.21 The associ-ation between breastfeeding and young age at diagnosis,found in our cohort, differs from the protective effect suggestedby other, limited data on the subject in pediatric IBD.22 Therelevance of the correlation between breastfeeding and L2remains to be confirmed. The breastfed group was heteroge-neous and recall bias is often a problem in studies recordingbreastfeeding.23 Therefore these data should be interpretedwith extreme caution.

In 9.2% of patients the diagnosis of Crohn's disease waspreceded by an episode of self limiting diarrhea. It cannot bedetermined whether these episodes were of infectious orinflammatory origin. Most of them belonged to the youngerage category at diagnosis. This association is not surprising,as younger children are more prone to infections in general.

Some vaccination data was present in the majority ofpatients even though most of this data was incomplete.Only polio vaccination is mandatory. General vaccinationcoverage is known to be over 80% because all recommendedvaccinations are offered free of charge in special pediatricclinics.24,25 This was not reflected by our data,mostly becauseof missing and incomplete data. For CD patients additionalvaccinations were recently recommended to prevent oppor-tunistic infections.26–28 Adherence to these recommenda-tions is very low so far. We note that even though nationalvaccination coverage is excellent, the information is oftennot recorded in the medical files or not transferred whenpatients transition to adult care. Clearly, more attention is




Table 4 Associations between variables at diagnosis for the whole cohort: p values. NS: Non-significant, No trues = No patients received this treatment.

Variables at diagnosis Gestationalage

Birthweight

Mode ofdelivery

BF FamilyhistoryIBD

Family historyAuto-immunedisease

ABUse 3 mprior todiagnosis

Stressfulevents3 mprior todiagnosis

Infectiousepisodes3 m priortodiagnosis

Age atdiagnosis

z-scoresheight

Diseaseseverity

L1 L2 L3 L4

Age at diagnosis NS NS NS 0.0003 0.032 NS NS NS 0.015 NS NS NS NS NS NSDisease severity NS NS NS NS NS NS NS NS NS NS 0.004 0.013 NS 0.011 NSL1 NS NS NS NS NS NS NS NS NS NS NS 0.013L2 NS NS NS 0.014 NS NS NS NS NS NS NS NSL3 NS NS NS NS NS NS NS NS NS 0.057 NS 0.011L4 NS NS NS NS NS NS NS NS NS NS NS NSFamily history of IBD 0.032 NS NS NS NS NS NS5 ASA monotherapy NS NS 0.001 NS NS NS NSSteroids monotherapy 0.001 NS NS NS NS NS NSImmuno-modulatorsmonotherapy

NS 0.043 0.012 0.031 NS NS 0.039

biologicals monotherapy Notrues

Notrues

Notrues

Notrues

Notrues

Notrues

Notrues

Antibiotics monotherapy NS NS 0.021 NS NS NS NSEnteral nutritionmonotherapy

NS NS NS NS NS NS NS

Rectal monotherapy NS NS NS NS NS NS NS5 ASA_combinationtherapy

0.004 NS NS NS NS NS NS

Steroids_combinationtherapy

0.011 NS 0.045 NS NS NS 0.048

Immunomodulators_combinationtherapy


Biologicals_combinationtherapy


AB-combination therapy 0.002 NS NS 0.032 NS NS NSEnteral_combinationtherapy

NS NS 0.005 NS NS NS 0.043

Rectal_combinationtherapy


Significant results are in bold.

7

Pleasecite

thisarticle

as:DeGreefE,et

al,Profileofpediatric

Crohn'sdisease

inBelgium

,JCrohns

Colitis

(2013),http://dx.doi.org/10.1016/j.crohns.2013.04.016



Table 5 Comparison between variables at diagnosis and at inclusion in the database for previously diagnosed patients, expressed in p-values. NS = non-significant (i.e. p N 0.05),No Trues = no patient received this treatment as monotherapy at diagnosis.

Variables at diagnosis Diseasedurationuntilinclusion

BMI z-scoreat inclusion

Heightz-score atinclusion

Diseaseseverity atinclusion

Surgeryprior toinclusion

5 ASA combinationtherapy at inclusion

Steroidscombinationtherapy atinclusion

Immunomodulatorcombinationtherapy atinclusion

Biologicalscombinationtherapy atinclusion

Enteralnutritioncombinationtherapyat inclusion

Gestational age NS NS NS NS 0.024 NS NS NS NS NSSurgery prior to diagnosis NS NS NS NS NS NS NS NS NSFamily history for IBD NS NS NS NS NS 0.020 NS 0.004 NS NSAge at diagnosis b0.001 NS NS NS NS NS NS NS 0.086 NSZscore_height at diagnosis 0.008 0.025 b0.001 NS NS NS NS NS 0.074 NSZscore_bmi at diagnosis NS b0.001 NS NS NS NS NS NS NS NSDisease severity at diagnosis NS NS NS NS NS NS NS NS NS NSL1 NS NS NS 0.042 0.055 NS 0.039 NS 0.063 NSL2 0.053 NS NS NS NS NS NS 0.068 NS NSL3 NS NS NS 0.033 NS NS NS NS 0.089 NSStricturing disease at diagnosis NS NS NS NS 0.001 NS NS NS 0.050 NSConcomitant disease at diagnosis NS NS 0.001 NS NS NS NS NS 0.059 NS5 ASA monotherapy NS NS NS NS NSSteroids monotherapy NS NS NS NS NSImmunomodulatorsmonotherapy

NS NS 0.015 NS NS

Biologicals monotherapy No trues No trues No trues No trues No truesEnteral feeding monotherapy No trues No trues No trues No trues No trues5 ASA_combination therapy b0.001 NS NS NS NSSteroids_combination therapy NS NS NS 0.004 NSImmunomodulators_combinationtherapy

NS NS NS NS NS

Biologicals_combination therapy NS NS NS NS NSEnteral feeding_combinationtherapy

NS NS NS NS NS

Significant results are in bold.

8E.

DeGreef

etal.

Pleasecite

thisarticle

as:DeGreefE,et

al,Profileofpediatric

Crohn's

diseasein

Belgium,J

CrohnsColitis

(2013),http://dx.doi.org/10.1016/j.crohns.2013.04.016



Figure 2 Evolution of treatment between diagnosis and regis-tration in the database for previously diagnosed patients.

9

needed in documenting and updating vaccination status atdiagnosis knowing the risk of opportunistic infections in thispatient group due to the immunosuppressive medication aspart of their usual treatment.

Abdominal pain, diarrhea and weight loss were the predom-inant clinical symptoms as was reported in literature.14,16 A BMIz-score below −1 was noticed in half of our pediatric patients.The importance of growth failure as a presenting feature of CDstill needs to be emphasized. Early treatment and adequatenutrition are crucial for catch up growth and achieving fullheight and weight potential. The previously diagnosed groupdemonstrates symptom improvement following treatmentand a decrease in growth failure. Height z-scores were inverselycorrelated to disease severity meaning less severe diseaseimproved growth, reflected by better height z-scores. Therelationship between BMI z-scores at inclusion and heightz-scores at diagnosis, imply in general an even better weightgain compared to growth catch up. The median follow up of2.7 y in this group might not be enough to reach a decentgrowth catch up. In the newly diagnosed population however,patients with lower z-scores weremore likely to receive enteralsupplements as part of the initial therapy indicating a recent,more adequate therapeutic strategy for growth retardation atdiagnosis. Certain data link growth impairment to diseaselocation,16 this could not be confirmed in our cohort. The size ofthe study population or the limited group of patients withsevere growth retardation may influence this result.

BELCRO confirmed that children with CD present withextensive and severe disease and with high upper GI tractinvolvement. Compared to the Eurokids database, the rateof upper GI tract involvement is high (70% vs. 46.2%).29 Thiscan be due to the way upper GI tract involvement is defined.We looked at microscopy, macroscopy and imaging whilethe Eurokids database only looked at macroscopic findings.Isolated upper GI tract involvement was more alike 2% vs.4%.29 Details on the upper GI tract findings were not availableat this stage and it is possible that they weremainly not specificas involvement was based on endoscopic and/or radiologicdata. While evaluation of disease location for different agecategories, confirmed the finding of predominant colonicdisease in the children under 6 y of age,3,14,15,19 it has to bestated that this was a very small group and the differencewith ileal and ileo-colonic disease is small.

The majority of patients had mild to moderate disease atdiagnosis as evaluated by PCDAI and PGA. Even though ileal


and ileo-colonic disease seemed to be associated with diseaseseverity, multiple regression analysis only defined CRP as anindependent risk factor for disease severity. CRP is not part ofthe PCDAI as a reflection of disease severity but has provenuseful in several studies as a predictive factor for treatmentresponse and to reflect mucosal healing.30–32

Family IBD was more often found in young patients,possibly due to an earlier expression or recognition of thedisease in the genetically predisposed. Only a positive familyhistory for IBD was associated with initial treatment choicefor 5-ASA and rapid introduction of immunomodulators. Inthe previously diagnosed group, the use of 5-ASA was higherand patients on 5-ASA had a longer disease duration atinclusion so probably the use of 5-ASA in these patientsreflects treatment schedules before 2008, when this drugwas still frequently used for Crohn's colitis. The reason whyyounger patients or patients with a family history were morelikely to receive 5-ASA is unclear. Maybe because of thelesser side effects compared to steroids, but slow decreaseover time is visible.

Our data demonstrates the importance of adequatetherapy at diagnosis as more severe initial treatment resultsin less severe disease over time. Physicians in this cohort weredefinitely compelled to use steroids in combination therapy asinitial treatment for more severe disease. After a mean followup of 2.7 y for a subgroup, patients who presented with severedisease were more likely to receive steroids as initial treatmentand were also more likely to have extensive disease (L3), whileat inclusion, those patients appeared to be more controlledwith inactive PCDAI scores. The opposite is true for patientswith isolated ileal disease. While 65.1% of patients at inclusionhad received steroids as part of their previous treatment, only11.6% were still on this treatment. These findings are markedlybetter than the American data where, even though 61% ofchildren showed a response to steroid treatment at 1 y, 31%developed steroid dependency.33 Another study showed a 40%relapse within 18 months after discontinuation of this treat-ment.34 Because of relapse rates, important side effects ofsteroid treatment and to improve long term outcome,alternative maintenance therapy and even induction therapyare investigated such as early immunomodulator use,34,35

enteral therapy and in severe cases top down therapy withbiologicals.36 Enteral therapy, an effective treatment in thepediatric CD population, induces remission, improves growthand leads to mucosal healing.37–39 Despite its safety, our datademonstrate a lack of enteral therapy as an initial mono-therapeutic approach. The way this treatment is introducedby the medical team and the support to the patient and theparents determines its success. In the newly diagnosed patientswe notice that enteral supplements are often given to patientswith growth retardation indicating an increasing awareness ofthe importance of growth in pediatric CD patients. In this cohortdisease location, disease behavior and age at diagnosis did notinfluence treatment choice at diagnosis and disease outcome,except for patients with stricturing disease whom had a greaterneed for surgery during follow up.

Immunomodulators were frequently prescribed, while bi-ologicals only have a very limited place at diagnosis. Step downtherapy was used as initial treatment in 2 patients becauseof the extreme severe presentation of disease. A significantincrease (p b 0.001) in biological therapy and immunomodula-tors is noticed over time, reflecting the step up therapy





generally used in pediatrics. A decreasing use of steroids inpatients with perianal disease over time might reflect theemerging therapeutic alternatives, like biological, for thisdisease behavior. Today, except for the use of budesonide inmild to moderate ileal disease and the use of biologicals infistulizing disease,40 therapeutic strategy is not influenced bydisease location or behavior even though we notice in ourcohort that patients with extensive disease tend to receivesteroids more frequently at diagnosis and biologicals incombination therapy during follow up. Physicians are lessinclined to use biologicals in ileal disease in order to avoidstricture formation. Recent American data compared the use ofbiologicals and immunomodulators within the first 30 days afterdiagnosis in a diverse group of newly diagnosed CD patients. Nosignificant outcome differences were noted except thatinfliximab treated children tended to be sicker at diagnosis.41

The RISK study group looked prospectively at the patientswith deep ulcerations on colonoscopy at diagnosis. This wasassociated with worse clinical parameters at diagnosis andworse disease severity scores (PCDAI/PGA) at 1 y follow-up,more so for patients who lacked treatment with immuno-modulators or biologicals within 3 months after diagnosis.42

These findings indicate the possible important effect ofadequate initial treatment on long term outcome data.Tailoring treatment becomes the subject of multiple studies;therefore the study of natural disease history is an importantstarting point. BELCRO illustrates the management of pediat-ric CD in Belgium. Treatment trends change over time andclear pediatric guidelines often need an update.43 Actualmanagement is based on adult experience and expert opinion,even though children are considered as a separate group.43

Further follow up of the BELCRO cohort will help to provide abetter insight in the impact of therapeutic strategy on diseasecourse by comparing previously and newly diagnosed patients.Presenting features should help determine individualizedtherapeutic regimens in the future.

Acknowledgments

BELCRO is supported by a grant from Schering-Plough, anMSD company.

J. M. Mahachie and K. VanSteen acknowledge researchopportunities offered by the Belgian Network DYSCO(Dynamical Systems, Control, and Optimization) funded by theInteruniversity AttractionPoles Programme, initiated by theBelgianState, Science Policy Office. Their work was alsosupported inpart by the IST Programme of the EuropeanCommunity, under the PASCAL2 Network of Excellence(Pattern Analysis, Statistical Modelling and ComputationalLearning), IST-2007-216886. In addition, J.M. MahachieJohn acknowledges study grant from Fonds de la RechercheScientifique (R.FNRS.2464 – F), Belgium.

WA, OD, HP, FB, GD, JFR, IE, OB, FF, VM, FVM and JCCparticipated in the design and concept of the study and theacquisition of data. IH, BH, FS, SVB, MS, IP, PA, AVG, EL andSV participated in the design and concept of the study, theacquisition of data, the data interpretation and the criticalreview of the manuscript. JJMM and KVS participated in thedata interpretation and were responsible for the major partof the data analysis. They critically reviewed the manu-script. EDG and GV participated in the design and concept of


the study, the acquisition of data and the data analysis andinterpretation. EDG drafted the manuscript with the helpand critical review of GV. All authors read and approved thefinal manuscript.

References

1. Benchimol EI, Fortinsky KJ, Gozdyra P, Van den Heuvel M, VanLimbergen J, Griffiths AM. Epidemiology of pediatric inflammatorybowel disease: a systematic reviewof international trends. InflammBowel Dis 2010;17(1):423–39.

2. Kelsen J, Baldassano RN. Inflammatory bowel disease: thedifference between children and adults. Inflamm Bowel Dis2008;14(Suppl 2):S9–S11.

3. Van Limbergen J, Russell RK, Drummond HE, et al. Definition ofphenotypic characteristics of childhood-onset inflammatorybowel disease. Gastroenterology 2008;135(4):1114–22.

4. Cosnes J, Gower-Rousseau C, Seksik P, Cortot A. Epidemiologyand natural history of inflammatory bowel diseases. Gastroen-terology 2010;140(6):1785–94 e4.

5. Barrett JC, Hansoul S, Nicolae DL, et al. Genome-wide associationdefines more than 30 distinct susceptibility loci for Crohn'sdisease. Nat Genet 2008;40(8):955–62.

6. BaldeM. Natural history of Crohn's disease in children: a population-based cohort study in Northern France. J Pediatr GastroenterolNutr 2006;43:S 11[Suppl.].

7. Hildebrand H, Finkel Y, Grahnquist L, Lindholm J, Ekbom A, AsklingJ. Changing pattern of paediatric inflammatory bowel disease innorthern Stockholm 1990–2001. Gut 2003;52(10):1432–4.

8. Hyams JS, Ferry GD, Mandel FS, et al. Development and validationof a pediatric Crohn's disease activity index. J PediatrGastroenterolNutr 1991;12(4):439–47.

9. Silverberg MS, Satsangi J, Ahmad T, et al. Toward an integratedclinical, molecular and serological classification of inflammatorybowel disease: report of a Working Party of the 2005 MontrealWorld Congress of Gastroenterology. Can. J. Gastroenterol.2005;19(Suppl A):5–36.

10. Levine A, Griffiths A, Markowitz J, et al. Pediatric modification ofthe Montreal classification for inflammatory bowel disease: theParis classification. Inflamm Bowel Dis 2011;17(6):1314–21.

11. Inflammatory bowel disease in children and adolescents:recommendations for diagnosis — the Porto criteria. J PediatrGastroenterol Nutr 2005;41(1):1–7.

12. Recommended childhood and adolescent immunization schedules— United States, 2009. Pediatrics 2009;123(1):189–90.

13. Accessed athttp://www.zorg-en-gezondheid.be/Diseases/Vaccination/2012.

14. Castro M, Papadatou B, Baldassare M, et al. Inflammatory boweldisease in children and adolescents in Italy: data from the pediatricnational IBD register (1996–2003). Inflamm Bowel Dis 2008;14(9):1246–52.

15. Jakobsen C, Paerregaard A, Munkholm P, Wewer V. Paediatricinflammatory bowel disease during a 44-year period in CopenhagenCounty: occurrence, course and prognosis — a population-basedstudy from the Danish Crohn Colitis Database. Eur J GastroenterolHepatol 2009;21(11):1291–301.

16. Sawczenko A, Sandhu BK. Presenting features of inflammatorybowel disease in Great Britain and Ireland. Arch Dis Child2003;88(11):995–1000.

17. Turunen P, Kolho KL, Auvinen A, Iltanen S, Huhtala H, Ashorn M.Incidence of inflammatory bowel disease in Finnish children,1987–2003. Inflamm Bowel Dis 2006;12(8):677–83.

18. van der Zaag-Loonen HJ, Casparie M, Taminiau JA, Escher JC,Pereira RR, Derkx HH. The incidence of pediatric inflammatorybowel disease in the Netherlands: 1999–2001. J PediatrGastroenterol Nutr 2004;38(3):302–7.


http://www.zorg-en-gezondheid.be/Diseases/Vaccination/




11

19. Vernier-Massouille G, Balde M, Salleron J, et al. Natural historyof pediatric Crohn's disease: a population-based cohort study.Gastroenterology 2008;135(4):1106–13.

20. Pozler O, Maly J, Bonova O, et al. Incidence of Crohn disease inthe Czech Republic in the years 1990 to 2001 and assessment ofpediatric population with inflammatory bowel disease. J PediatrGastroenterol Nutr 2006;42(2):186–9.

21. Bolling K GC, Hamlyn B, Thornton A. Infant feeding survey 2005.In: The Information Centre GSS, ed. Leeds; 2007.

22. Barclay AR, Russell RK, Wilson ML, Gilmour WH, Satsangi J,Wilson DC. Systematic review: the role of breastfeeding in thedevelopment of pediatric inflammatory bowel disease. J Pediatr2009;155(3):421–6.

23. Bland RM, Rollins NC, Solarsh G, Van den Broeck J, Coovadia HM.Maternal recall of exclusive breast feeding duration. Arch DisChild 2003;88(9):778–83.

24. Vellinga A, Depoorter AM, Van Damme P. Vaccination coverageestimates by EPI cluster sampling survey of children (18–24 months)in Flanders, Belgium. Belgium. Acta Paediatr 2002;91(5):599–603.

25. http://www.zorg-en-gezondheid.be/Diseases/Vaccination/2012.26. Veereman-Wauters G, de Ridder L, Veres G, et al. Risk of infection

and prevention in pediatric patients with IBD: ESPGHAN IBD PortoGroup Commentary. J Pediatr Gastroenterol Nutr 2012;54(6):830–7.

27. De Greef E, Vandenplas Y, Veereman-Wauters G. Opportunisticinfections in paediatric inflammatory bowel disease patients. ArchDis Child 2012;97(1):5–7.

28. Rahier JF, Ben-Horin S, Chowers Y, et al. European evidence-basedconsensus on the prevention, diagnosis and management ofopportunistic infections in inflammatory bowel disease. J CrohnsColitis 2009;3:47–91.

29. de Bie CI, Buderus S, Sandhu BK, et al. Diagnostic workup ofpaediatric patients with inflammatory bowel disease in Europe:results of a 5-year audit of the EUROKIDS registry. J PediatrGastroenterol Nutr 2012;54(3):374–80.

30. Costantino G, Furfaro F, Belvedere A, Alibrandi A, Fries W.Thiopurine treatment in inflammatory bowel disease: Responsepredictors, safety, and withdrawal in follow-up. J Crohns Colitis2010;6(5):588–96.

31. Daperno M, Castiglione F, de Ridder L, et al. Results of the 2ndpart Scientific Workshop of the ECCO. II: Measures and markers ofprediction to achieve, detect, and monitor intestinal healing ininflammatory bowel disease. J Crohns Colitis 2010;5(5):484–98.


32. Reinisch W, Wang Y, Oddens BJ, Link R. C-reactive protein, anindicator for maintained response or remission to infliximab inpatients with Crohn's disease: a post-hoc analysis from ACCENTI. Aliment Pharmacol Ther 2012;35(5):568–76.

33. Markowitz J, Hyams J, Mack D, et al. Corticosteroid therapy inthe age of infliximab: acute and 1-year outcomes in newlydiagnosed children with Crohn's disease. Clin GastroenterolHepatol 2006;4(9):1124–9.

34. Markowitz J, Grancher K, KohnN, Lesser M, DaumF. Amulticentertrial of 6-mercaptopurine and prednisone in children withnewly diagnosed Crohn's disease. Gastroenterology 2000;119(4):895–902.

35. Punati J, Markowitz J, Lerer T, et al. Effect of early immunomod-ulator use inmoderate to severe pediatric Crohn disease. InflammBowel Dis 2008;14(7):949–54.

36. Rubin DT, Uluscu O, Sederman R. Response to biologic therapyin Crohn's disease is improved with early treatment: An analysisof health claims data. Inflamm. Bowel Dis. 2010.

37. Borrelli O, Cordischi L, Cirulli M, et al. Polymeric diet aloneversus corticosteroids in the treatment of active pediatricCrohn's disease: a randomized controlled open-label trial. ClinGastroenterol Hepatol 2006;4(6):744–53.

38. Kirschner BS, Klich JR, Kalman SS, deFavaro MV, Rosenberg IH.Reversal of growth retardation in Crohn's disease with therapyemphasizing oral nutritional restitution. Gastroenterology1981;80(1):10–5.

39. Papadopoulou A, Holden CE, Paul L, Sexton E, Booth IW. Thenutritional response to home enteral nutrition in childhood.Acta Paediatr 1995;84(5):528–31.

40. McKeage K, Goa KL. Budesonide (Entocort EC Capsules): a reviewof its therapeutic use in themanagement of active Crohn's diseasein adults. Drugs 2002;62(15):2263–82.

41. Markowitz JEJ, Pfefferkorn MD, et al. Infliximab versus immuno-modulator as first maintenance therapy in children with Crohndisease. Gastroenterology 2012;142(5):S 349.

42. Hyams JKM, Denson L, et al. Effect of deep ulceration atdiagnostic colonoscopy on initial therapy and one year outcomein children with Crohn's disease. Gastroenterology 2012;142(5):S36.

43. Van Assche G, Dignass A, Reinisch W, et al. The second Europeanevidence-based Consensus on the diagnosis and management ofCrohn's disease: Special situations. J Crohns Colitis 2011;4(1):63–101.





Documents

Profile of pediatric Crohn's disease in Belgium