CARD15 Gene Mutations and Risk for Early Surgery in Pediatric-Onset Crohn s Disease

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2004;2: CARD15 Gene Mutations and Risk for Early Surgery in Pediatric-Onset Crohn s Disease SUBRA KUGATHASAN,* NICOLE COLLINS,* KAREN MARESSO, RAYMOND G. HOFFMANN, MICHAEL STEPHENS,* STEVEN L. WERLIN,* COLIN RUDOLPH,* and ULRICH BROECKEL, *Department of Pediatrics, Department of Medicine, Human Molecular Genetics Center, and the Health Policy Institute & Biostatistics, Medical College of Wisconsin, Milwaukee, Wisconsin Background & Aims: The risk for Crohn s disease (CD) is determined in part by genetic factors. Three recently described mutations in the CARD15(NOD2) gene have been associated with adult-onset CD. We investigated the effect of CARD15 mutations on disease manifestation, disease progression, and the risk for early surgery in childhood-onset CD. Methods: Genotyping for 3 CARD15 mutations: R702W, G908R, and 3020insC, was performed in 186 children with CD from a prospective cohort. A transmission-disequilibrium test was used to test for association with CD. Genotype with disease location and behavior was tested with logistic regression analysis. The effect of mutations on surgical outcome was evaluated using a Cox proportional hazard analysis. Results: The mean age at CD diagnosis was 12.4 years. The frequency of allelic mutations observed was 6.6% for R702W, 6% for G908R, and 13.1% for 3020insC. Of Caucasian CD children, 42% had at least one CARD15 mutation. None of the non-caucasian children with CD had any CARD15 mutation. A significant association was detected for 3020insC (P.0045). Ileal location (odds ratio, 4.3; P.003) and stricturing disease (odds ratio, 6.6; P.0001) was more frequent and the risk for surgery was higher (hazard ratio, 5.8; P <.0001) and surgery occurred earlier (hazard ratio, 2.24) in those children with 3020insC mutation compared with those without 3020insC. Conclusions: In children with pediatric-onset CD, early development of stricturing behavior leading to surgical resection is influenced by ileal location and 3020insC variant of the CARD15 mutation. Genetic testing may identify children with CD who are at risk for early surgery. The pathogenesis of Crohn s disease (CD) entails a complex interaction between environmental factors, immunologic influence, and a genetic predisposition. 1,2 An inflammatory bowel disease (IBD) susceptibility locus (IBD1) was identified on chromosome 16 by using genome-wide genetic linkage analysis 3 and, subsequently, functional variants were shown to be risk factors for CD. 4,5 Three major coding region polymorphisms (R702W, G908R, and 3020insC) within the CARD15 gene (also known as NOD2) have been associated with CD among adult patients of European descent. Subsequently, studies in adults have examined the impact of these mutations on age of onset, site of disease involvement, disease behavior, disease progression, and response to treatment A strong association between CARD15 mutations and ileal disease location has been shown in adults 6 10 and in 2 small pediatric studies. 12,13 Among adult patients, stricturing (fibrostenosing) phenotype appears to correlate with CARD15 variants, 7,14 but some investigators dispute this correlation because the CARD15 genotype was not identified uniformly as an independent factor. 9,11 We studied genotype-phenotype interactions in a cohort of children with CD to determine if the frequency of known genetic variants associated with CD was increased. The power of genetic studies exploring the role of genetic factors on disease presentation may increase when focused on patients with early onset disease in whom it is likely that there is a higher gene dosage effect 15 and less environmental influence from factors such as smoking. 16 Furthermore, in a disorder such as CD, different gene variants may be associated with divergent disease behaviors with regard to age of presentation, location, and severity of disease. Materials and Methods Human Subjects All children attending the IBD clinic at the Children s Hospital of Wisconsin during a 1-year period were asked to participate in this study. Of 255 patients, 98% (n 250) agreed. Blood collection for DNA extraction consisted of trios of father, mother, and affected child. All the children were or previously had been phenotyped carefully from the time of Abbreviation used in this paper: CD, Crohn s disease; CI, confidence interval; IBD, inflammatory bowel disease; UC, ulcerative colitis by the American Gastroenterological Association /04/$30.00 PII: /S (04)

2 1004 KUGATHASAN ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 2, No. 11 diagnosis, and phenotype was followed-up prospectively as a part of a statewide population-based pediatric IBD registry. 17 Observations recorded include the site of disease, involvement at the time of diagnosis and during the clinical follow-up evaluation, family history, radiologic and endoscopic diagnostic procedures, histologic findings, medical therapies, and surgical interventions. The study was reviewed and approved for human subject participation by the Children s Hospital of Wisconsin Institutional Review Board. Familial occurrence of disease was defined as the presence of IBD in a first- or second-degree relative. Positive histories were verified by contacting the actual relative. Medication history including the number of corticosteroid treatment courses, 5-aminosalicylic acid, immunomodulators (azathioprine, 6-mercaptopurine, or methotrexate), and biologic therapies (infliximab) also were extracted from the prospective database. Criteria for Inflammatory Bowel Disease Diagnosis and Disease Subcategorization The diagnosis of IBD required endoscopic evaluation, including histologic assessment of mucosal pinch biopsy specimens. All patients underwent colonoscopic examination with multiple biopsy examinations and most had upper endoscopy performed. Stool studies were performed in all patients to exclude parasitic and bacterial causes. In all cases, small-bowel follow-through barium radiographs and/or colonoscopy and ileoscopy evaluated the terminal ileum. In all patients, the diagnosis of IBD was confirmed by a clinical pathologist and pediatric conditions manifesting with chronic diarrhea including allergic colitis and eosinophilic gastroenteropathy were excluded. Serologic testing with the identification of antineutrophil and/or anti-saccharomyces cerevisiae antibodies were not considered sufficient criteria for a formal diagnosis of IBD and were obtained in only 10% of patients. No attempts were made for phenotype or serologic correlation with genotyping. All IBD cases were subcategorized further into 3 clinical groups CD, ulcerative colitis (UC), and indeterminate colitis defined according to established clinical, biochemical, radiologic, endoscopic, and histologic criteria. CD was defined as the presence of histologically confirmed discontinuous chronic inflammation of the gastrointestinal tract confirmed by endoscopy, and supported by clinical, biochemical, and radiologic evidence of CD. UC was defined as the presence of continuous inflammation limited to the colon, extending proximally from the involved rectum, with histologically typical chronic inflammation limited to the mucosa, not extending beyond the muscularis mucosa. When IBD was present but a definitive diagnosis of CD or UC could not be made because of overlapping features, these patients (n 2) were excluded and genotyping was not performed. Determination of Disease Site and Extent In each CD patient, the anatomic distribution of the disease was determined by involvement of regions of CD as orofacial (ie, orofacial granulomatosis), upper-tract involvement (esophageal, gastric, duodenal), small bowel (jejunal, ileal), colonic, and perianal. With the exception of jejunal and perianal CD, all other areas of anatomic involvement were confirmed by microscopic evidence of CD chronic inflammation. The presence of esophagitis, gastritis, or duodenitis characterized as nonspecific inflammation (ie, acid peptic disease, and so forth) was not sufficient criteria for the diagnosis of upper gastrointestinal tract CD. Patients were categorized further into 1 of 5 groups based on the maximal extent of the disease: (1) ileal: gross involvement of the ileum without colonic disease but could include upper-track involvement; (2) ileocolonic: gross involvement of the ileum and colon but could include any upper-tract involvement; (3) colonic: gross involvement of the colon without small-bowel involvement; (4) perianal disease: perianal fistulas, perianal abscesses, but not perianal fissures or perianal tags (this was considered a separate category and could occur with any other site of involvement); and (5) upper gastrointestinal tract involvement: evidence of gross or microscopic involvement of the upper gastrointestinal tract (this could include any of the first 4 categories). Disease Behavior Patients were categorized based on the clinical history with either fistulizing, stricturing, or inflammatory disease behavior by using Vienna classification guidelines criteria. 18 Fistulizing or penetrating disease was defined as the presence of intra-abdominal fistulizing lesions (the presence of enteroenteric fistulas) or perianal fistulizing lesions (enterocutaneous). When significant strictures were present that required surgery in the absence of fistulizing disease, they were classified into the stricturing (fibrostenosing) phenotype. This stringent criterion of including only stricture patients in this category was used to prevent a subset of patients who might have ileal narrowing from inflammation, not stricturing, on radiologic examination. Patients with nonstricturing and nonfistulizing disease at the time of presentation and throughout follow-up to date were classified as having inflammatory disease behavior. Surgery for Crohn s Disease All patients who underwent surgery (n 29, 18% of our CD cohort) had either ileocolonic or ileal resection after failing medical therapy. Indications for surgery included stricturing disease confirmed by radiologic examination (ileal narrowing by small-bowel follow-through) in all patients. In addition to ileal narrowing, localized abscess formation with an inflammatory mass was seen in 8% (n 3) of the patients, and growth failure was seen in 58% (n 17) of the patients at the time they underwent surgery. CARD 15 (NOD2) Genotyping All patients and available parents were genotyped for each of the 3 CARD15 variants. The R702W single nuclear polymorphism was genotyped by using direct fluorescencebased sequencing (ABI 3700 automated sequencer) and big- Dye chemistry (Applied Biosystems) using forward (5=-CA- CAACCTTCAGATCACAGCAG-3=) and reverse (5=-

3 November 2004 CARD15 GENE MUTATIONS 1005 Table 1. Clinical Characteristics of 163 Caucasian Children With CD Characteristic No. (%) of patients (N 163) Male 94 (57.7) Mean age at diagnosis (range) 12.4 (3 18) Positive family history 32 (19.6) Disease localization Small bowel only 32 (19.6) Colon only 46 (28.2) Small bowel and colon 84 (51.5) Upper gastrointestinal tract 70 (42.9) (including microscopic) Perianal 59 (36.2) Disease behavior Inflammatory 104 (64) Perforating 29 (18) Stricturing 29 (18) ATGCCCAGTAACACTCACTACAGA-3=). Both the G908R and the 3020insC were genotyped using ABI s Taqman Assays-by-Design technology on a ABI 7900HT sequence detection system (ABI Applied BioSystems, Foster City, CA). Statistical Analysis Allele frequencies and Hardy Weinberg equilibrium were calculated in patients with CD and UC. The transmission-disequilibrium test was used to assess linkage and association within the trios (GeneHunter software, version 2.0). Both univariate and multivariate (controlling for age at diagnosis and sex) logistic regression analyses were performed for discrete phenotypes to determine the correlation with CARD15 mutational status (the presence or absence of any of the 3 mutations), and, more specifically, with the presence or absence of the 3020insC variant. The probability of the need for surgery was estimated with the Kaplan Meier technique. For the analysis of prognostic factors (surgery-free survival), we used a Cox proportional-hazards regression model, stratified according to the 2 groups (0 vs. 1 or 2 copies of 3020insC). A test of the proportionality assumption was performed using the Schoenfeld residuals. Results Study Population and Prevalence of CARD15/NOD2 A total of 250 children with IBD was included in this study. These IBD children were part of a populationbased prospective IBD registry at the Children s Hospital of Wisconsin. Of these, 186 (74%) had CD, 62 (25%) had UC, and 2 had indeterminate colitis (1%). Allele frequencies were calculated in racial and diagnostic groupings. Although the number of non-caucasian children was small, none of the 3 CARD15(NOD2) mutations were identified within this subset (20 African American and 6 Hispanic). Additionally, neither the R702W nor the C3020ins was identified within the UC group, although the G908R was observed at a frequency of 2.3%. Based on these findings, we excluded non- Caucasian children with CD and those diagnosed with UC or indeterminate colitis from further analysis. Among the 163 Caucasian CD patients, the mean age at diagnosis was 12.4 years (range, 3 18 y), with an average follow-up time of 39 months (range, 6 88 mo) after diagnosis (Table 1). Allele frequencies for R702W, G908R, and C3020insC were 6.6%, 6.0%, and 13.1%, respectively (Table 2). Forty-two percent of children were found to have at least one CARD15 mutation, but only 7% had 2 mutations. A family history of IBD was present among 24% of CARD15-positive patients and in 18% of those with no allelic variant. Only 4% of older adolescents gave a history of current smoking or past history of smoking. Association Analysis Sixty-one Caucasian CD trios were available for association testing using the transmission-disequilibrium test. A significant association with CD was detected for the 3020insC polymorphism (P.0045). We did not detect a significant association with the other polymorphisms. Genotype-Phenotype Correlation We performed a logistic regression analysis to determine the effect of CARD15 and 3020insC status on disease localization, disease behavior, and treatments (steroids, infliximab, and so forth). Phenotypic frequencies also were calculated and stratified by genotypic status. An unpaired t test was used to compare the mean age at diagnosis between the groupings. Results are summarized in Table 3. CARD15 status confers a 4.3- fold increased risk (age-adjusted; 95% confidence interval [CI]: ; P.003) of small-bowel localization in children carrying at least 1 of the 3 variants, as compared with children with none of the variants. Twenty-nine percent of patients with at least one CARD15 variant had small-bowel localization compared with 11% Table 2. Genotypic and Allelic Frequencies of the 3 CARD15 Disease Predisposing Mutations in 163 Caucasian Children With CD Genotype R702W G908R 3020insC / / / Allele frequency

4 1006 KUGATHASAN ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 2, No. 11 Table 3. Relationship Between CARD15 Variants and Disease Phenotype Univariate analysis CARD15 CARD insC 3020insC CARD insC Characteristics N 58 N 80 N 36 N 113 OR 95% CI P OR 95% CI P Age at diagnosis (mean SE) Disease location % Ileum a a Ileum and colon Colon only a a Perianal disease Upper tract (including microscopic) Disease behavior % Inflammatory a Fistulizing (penetrating) Stricturing (fibrostenosing) Family history of IBD NOTE. Results of disease location, disease behavior, and family history by genotype. The results are shown by presence (CARD15 ) or absence (CARD15 ) of all 3 mutations as well as the most frequently seen 3020insC variant. OR, odds ratio. a Age-adjusted P values. of CARD15-negative patients. The 3020insC showed a similar effect as CARD15 (age-adjusted odds ratio, 4.52; 95% CI: ; P.002). Thirty-six percent of those with the 3020insC had small-bowel disease, whereas only 12% of 3020insC-negative patients had small-bowel disease. In addition, CARD15 status was shown to significantly protect against colonic involvement (negative correlation). Significant genotype-phenotype correlations could not be identified for ileocolonic, perianal, upper gastrointestinal tract (including microscopic involvement), or orofacial localization. Similarly, no correlations were observed among CARD15-positive and -negative patients and the need for infliximab or corticosteroid dependency. Disease Behavior At presentation, the vast majority (90%) had inflammatory behavior, 3% with stricturing behavior and 8% with fistulas, in our cohort of 163 patients. The CD children in this cohort were followed-up for a mean period of 3.2 years. Disease type was classified as inflammatory (nonstricturing, nonpenetrating) in 105 (64%) patients, stricturing in 29 (18%) patients, and fistulizing in 29 (18%) patients during the most recent follow-up evaluation. Among all of the Caucasian children with CD, only 3 (2%) had strictures requiring surgery at the time of diagnosis, compared with 29 (18%) who required surgery for strictures overall during follow-up evaluation. Both CARD15 and 3020insC status showed highly significant associations with stricturing complications. The odds of stricturing in children carrying at least one allele of the 3020insC were 6.6-fold (odds ratio, 6.62; 95% CI: ; P.0001) compared with those of children not carrying this variant. Within our study population, 42% of children positive for the 3020insC had stricturing, whereas only 10% of those without the 3020insC had stricturing. Given this association of CARD15 variants with a more severe disease phenotype, both CARD15 and the 3020insC were significantly protective (negative correlation) against inflammatory disease behavior (CARD15: age-adjusted odds ratio,.34; 95% CI:.16.71; P.005; 3020insC: odds ratio,.43; 95% CI:.19.93; P.03). Significant associations could not be identified for fistulizing disease. Age at Diagnosis There was no difference in the mean age at diagnosis between Caucasian children with CD with at least one of the CARD15 variants and those without any of the variants (11.98 vs ; P.40). Similarly, there was no difference in age at diagnosis between children with CD with the 3020insC or without the 3020insC variant (12.46 v ; P.81).

5 November 2004 CARD15 GENE MUTATIONS 1007 Figure 1. Kaplan Meier analysis showing patients without surgery according to the time between the diagnosis and surgery in CD without 3020insC (solid line, n 112), and children with CD who carry at least one allele of 3020insC variant (interrupted line). There was a significant risk for surgery (P.0001; hazard ratio, 5.8) in children with CD who carry 3020insC compared with children with CD without 3020insC during the follow-up evaluation. Survival Analysis Stricturing is an especially clinically relevant phenotype given its subsequent complications and need for surgical treatment. Given the strong association of the 3020insC variant with stricturing phenotype, we used Cox proportional hazards modeling to determine the effect of this single nuclear polymorphism on time to surgery. Our aim was to determine whether 3020insC mutation testing could characterize a high-risk subgroup of Caucasian patients with CD who would benefit from more aggressive early therapy, possibly leading to better outcomes. Time to surgery was calculated in months for all patients. The Cox proportional hazard assumption was determined to be satisfied by comparing observed vs. predicted survival curves. We first evaluated time to surgery among all Caucasian children with CD stratified by the presence or absence of the 3020insC mutation. The Cox proportional hazards assumption was tested with the Schoenfeld residuals (P.35). The unadjusted hazard ratio was 6.07 (95% CI: ; P.0001), indicating that the hazard for surgery is 6 times greater in children with at least one copy of the 3020insC allele than in those children without this variant (Figure 1). The hazard ratio changed little when adjusted for small-bowel location and disease behavior at diagnosis (5.83; 95% CI: ; P.0001). When we analyzed the risk for surgery with a pure control group (children with CD without any CARD15 mutations), the risk was much higher in children with CD with 3020insC (adjusted hazard ratio, 7.78; 95% CI: ; P.0005). In the subset of children who underwent surgery for stricturing complications (n 29), those with a 3020insC mutation showed a trend toward a need for earlier surgery (hazard ratio, 2.61; 95% CI: ; P.048). The median time to surgery in those children with the 3020insC variant (n 11) was 14 months, whereas the median time to surgery in those children without a CARD15 mutation (n 15) was 23 months. These results show that children positive for the 3020insC variant are at a significantly higher risk for the need for surgery than children without the 3020insC mutation, suggesting that this variant could act as a prognostic factor for early surgery in Caucasian children with CD. Discussion Previous studies in adult patients with CD showed an association with specific mutations of the CARD 15 gene In our pediatric CD cohort with a mean age of disease onset at 12.4 years, we found that 3020insC confers strong susceptibility to CD in Caucasian children. In contrast, the frequency of 2 other mutations known to be associated with adult-onset CD (R702W and G908R) failed to reach significant association. Furthermore, when the disease duration was taken into account, stricturing complications leading to early surgery were found more frequently in children with CD with the 3020insC mutation compared with those children with CD without CARD15 mutations. In our cohort, 42% of the children with CD had at least one CARD15 (NOD2) mutation. This is similar to

6 1008 KUGATHASAN ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 2, No. 11 the overall frequency described in many studies involving a Caucasian adult population. 4,8 10,14,19,20 However, comparing the frequency of each tested individual mutation from our study with adult patients with CD showed selected differences. Although the frequency in our pediatric cohort for the 3020insC was 13% and is within the range of 6% 16% seen in adults, we observed a decreased frequency of 6.6% for the R702W mutation compared with 9% 17% in an adult Caucasian population with CD. The allele frequency for the G908R was similar between our study and the adult studies. With regard to disease association and CARD15 mutations, we observed a significant transmission-disequilibrium test for the 3020insC, whereas the other 2 polymorphisms did not reach statistical significance. Although the 3020insC showed the strongest association, because of our sample size and the decreased allele frequencies of R702W and G908R, we cannot exclude these 2 lessfrequent polymorphisms from contributing to CD in children. Similar to adult studies, we did not observe a strong association for UC with CARD15 mutations. Because we did not observe a substantial allele frequency difference between pediatric-onset and adult-onset CD patients, this might indicate that the age of onset in those with CD is unlikely to be influenced significantly by CARD15 (NOD2). The association between CARD15 and disease behavior is controversial. Some studies have shown an association between CARD15 variant and a stricturing phenotype, 14,20 but this has not been observed in all studies It has been shown that the majority of CD patients present with nonstricturing disease at diagnosis, whereas after several decades of disease presence a significant proportion harbor a stricturing pattern 21,22 that often requires surgical resection. This changing behavior may be determined by both genetic and environmental factors. Longer duration of disease and smoking are associated with a more aggressive CD course. 23,24 The mean age of onset in our CD cohort was 12.4 years, with a mean of 3.2 years follow-up evaluation; the percentage of our CD cohort ever exposed to smoking was less than 5%, smoking was an unlikely factor that would have influenced the disease behavior in our patients. A subset of pediatric-onset CD patients may have very aggressive disease manifesting with stricturing disease and growth failure leading to early surgical resection An attempt to identify markers or factors that are associated with these phenotypes is greatly needed because predicting this behavior may allow clinicians to select patients for early aggressive medical management in the hopes of preventing surgery. Two retrospective, multicenter, chart-review studies did not detect a significant effect of CARD15 variants on the need for surgery in adult-onset CD. 11,20 In contrast, we found that children with the 3020insC have a 6.6-fold increased risk for developing a stricturing phenotype requiring surgery. This difference may be explained by the inherent limitations of a retrospective study design compared with our prospective design or may be owing to a more aggressive disease pattern in our children with CD with 3020insC variants compared with other patients so that they progress to surgery at a younger age in this childhoodonset CD cohort. In general, a prospective study design should provide the most compelling evaluation for any disease marker with regard to outcomes. In conclusion, the 3020insC mutation within the CARD15 (NOD2) gene appears to confer susceptibility to CD in Caucasian children from North America. In children with CD, the 3020insC mutation increases the likelihood of ileal and stricturing (fibrostenotic) disease that will require early surgical resection. It is possible that genotyping at the time of diagnosis will identify a subgroup of CD children who are at risk for a more rapid development of complications (stricturing). We speculate that these patients may benefit from the early use of more aggressive therapies such as immunomodulators and biological therapies. Although the findings presented here are specific to our cohort of CD children, we believe this observation represents a proof-of-principal on how gene identification strategies combined with genotype/phenotype correlations might identify prognostic indicators to guide disease management. References 1. Fiocchi C. Inflammatory bowel disease: etiology and pathogenesis. Gastroenterology 1998;115: Oliva-Hemker M, Fiocchi C. Etiopathogenesis of inflammatory bowel disease: the importance of the pediatric perspective. Inflamm Bowel Dis 2002;8: Hugot JP, Laurent-Puig P, Gower-Rousseau C, Olson JM, Lee JC, Beaugerie L, Naom I, Dupas JL, Van Gossum A, Orholm M, Bonaiti-Pellie C, Weissenbach J, Mathew CG, Lennard-Jones JE, Cortot A, Colombel JF, Thomas G. Mapping of a susceptibility locus for Crohn s disease on chromosome 16. Nature 1996;379: Hugot JP, Chamaillard M, Zouali H, Lesage S, Cezard JP, Belaiche J, Almer S, Tysk C, O Morain CA, Gassull M, Binder V, Finkel Y, Cortot A, Modigliani R, Laurent-Puig P, Gower-Rousseau C, Macry J, Colombel JF, Sahbatou M, Thomas G. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn s disease. Nature 2001;411: Ogura Y, Bonen DK, Inohara N, Nicolae DL, Chen FF, Ramos R, Britton H, Moran T, Karaliuskas R, Duerr RH, Achkar JP, Brant SR, Bayless TM, Kirschner BS, Hanauer SB, Nunez G, Cho JH. A frameshift mutation in NOD2 associated with susceptibility to Crohn s disease. Nature 2001;411: Abreu MT. The pathogenesis of inflammatory bowel disease:

7 November 2004 CARD15 GENE MUTATIONS 1009 translational implications for clinicians. Curr Gastroenterol Rep 2002;4: Hampe J, Grebe J, Nikolaus S, Solberg C, Croucher PJ, Mascheretti S, Jahnsen J, Moum B, Klump B, Krawczak M, Mirza MM, Foelsch UR, Vatn M, Schreiber S. Association of NOD2 (CARD 15) genotype with clinical course of Crohn s disease: a cohort study. Lancet 2002;359: Lesage S, Zouali H, Cezard JP, Colombel JF, Belaiche J, Almer S, Tysk C, O Morain C, Gassull M, Binder V, Finkel Y, Modigliani R, Gower-Rousseau C, Macry J, Merlin F, Chamaillard M, Jannot AS, Thomas G, Hugot JP. CARD15/NOD2 mutational analysis and genotype-phenotype correlation in 612 patients with inflammatory bowel disease. Am J Hum Genet 2002;70: Ahmad T, Armuzzi A, Bunce M, Mulcahy-Hawes K, Marshall SE, Orchard TR, Crawshaw J, Large O, de Silva A, Cook JT, Barnardo M, Cullen S, Welsh KI, Jewell DP. The molecular classification of the clinical manifestations of Crohn s disease. Gastroenterology 2002;122: Cuthbert AP, Fisher SA, Mirza MM, King K, Hampe J, Croucher PJ, Mascheretti S, Sanderson J, Forbes A, Mansfield J, Schreiber S, Lewis CM, Mathew CG. The contribution of NOD2 gene mutations to the risk and site of disease in inflammatory bowel disease. Gastroenterology 2002;122: Louis E, Michel V, Hugot JP, Reenaers C, Fontaine F, Delforge M, El Yafi F, Colombel JF, Belaiche J. Early development of stricturing or penetrating pattern in Crohn s disease is influenced by disease location, number of flares, and smoking but not by NOD2/ CARD15 genotype. Gut 2003;52: Sun L, Roesler J, Rosen-Wolff A, Winkler U, Koch R, Thurigen A, Henker J. CARD15 genotype and phenotype analysis in 55 pediatric patients with Crohn disease from Saxony, Germany. J Pediatr Gastroenterol Nutr 2003;37: Tomer G, Ceballos C, Concepcion E, Benkov KJ. NOD2/CARD15 variants are associated with lower weight at diagnosis in children with Crohn s disease. Am J Gastroenterol 2003;98: Abreu MT, Taylor KD, Lin YC, Hang T, Gaiennie J, Landers CJ, Vasiliauskas EA, Kam LY, Rojany M, Papadakis KA, Rotter JI, Targan SR, Yang H. Mutations in NOD2 are associated with fibrostenosing disease in patients with Crohn s disease. Gastroenterology 2002;123: Lander ES, Schork NJ. Genetic dissection of complex traits. Science 1994;265: Cosnes J, Beaugerie L, Carbonnel F, Gendre JP. Smoking cessation and the course of Crohn s disease: an intervention study. Gastroenterology 2001;120: Kugathasan S, Judd R, Hoffman RG, Heikenen J, Telega G, Khan F, Weisdorf-Schindele S, San Pablo W Jr, Perroult J, Park R, Yaffe M, Brown C, Rivera-Bennett MT, Halabi I, Martinez A, Blank E, Werlin SL, Rudolph C, Binion DG, for the Wisconsin Pediatric IBD Alliance. Epidemiologic and clinical characteristics of children with newly diagnosed inflammatory bowel disease in Wisconsin: a statewide population-based study. J Pediatr 2003;143: Gasche C, Scholmerich J, Brynskov J, D Haens G, Hanauer SB, Irvine EJ, Jewell DP, Rachmilewitz D, Sachar DB, Sandborn WJ, Sutherland LR. A simple classification of Crohn s disease: report of the Working Party for the World Congresses of Gastroenterology, Vienna Inflamm Bowel Dis 2000;6: Vermeire S, Wild G, Kocher K, Cousineau J, Dufresne L, Bitton A, Langelier D, Pare P, Lapointe G, Cohen A, Daly MJ, Rioux JD. CARD15 genetic variation in a Quebec population: prevalence, genotype-phenotype relationship, and haplotype structure. Am J Hum Genet 2002;71: Brant SR, Picco MF, Achkar JP, Bayless TM, Kane SV, Brzezinski A, Nouvet FJ, Bonen D, Karban A, Dassopoulos T, Karaliukas R, Beaty TH, Hanauer SB, Duerr RH, Cho JH. Defining complex contributions of NOD2/CARD15 gene mutations, age at onset, and tobacco use on Crohn s disease phenotypes. Inflamm Bowel Dis 2003;9: Louis E, Collard A, Oger AF, Degroote E, Aboul Nasr El Yafi FA, Belaiche J. Behaviour of Crohn s disease according to the Vienna classification: changing pattern over the course of the disease. Gut 2001;49: Cosnes J, Cattan S, Blain A, Beaugerie L, Carbonnel F, Parc R, Gendre JP. Long-term evolution of disease behavior of Crohn s disease. Inflamm Bowel Dis 2002;8: Sutherland LR, Ramcharan S, Bryant H, Fick G. Effect of cigarette smoking on recurrence of Crohn s disease. Gastroenterology 1990;98: Cosnes J, Carbonnel F, Beaugerie L, Le Quintrec Y, Gendre JP. Effects of cigarette smoking on the long-term course of Crohn s disease. Gastroenterology 1996;110: Griffiths AM, Nguyen P, Smith C, MacMillan JH, Sherman PM. Growth and clinical course of children with Crohn s disease. Gut 1993;34: Griffiths AM, Wesson DE, Shandling B, Corey M, Sherman PM. Factors influencing postoperative recurrence of Crohn s disease in childhood. Gut 1991;32: Baldassano RN, Han PD, Jeshion WC, Berlin JA, Piccoli DA, Lautenbach E, Mick R, Lichtenstein GR. Pediatric Crohn s disease: risk factors for postoperative recurrence. Am J Gastroenterol 2001;96: Address requests for reprints to: Subra Kugathasan, MD, Department of Pediatrics, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, Wisconsin skuga@mcw.edu; fax: (414) Supported by National Institutes of Health grant (RR016111, to S.K.), Crohn s & Colitis Foundation of America, the General Clinical Research Center grant M01-RR00058 from the National Institutes of Health, the Children s Hospital Research Institute, and the Medical College of Wisconsin s Digestive Disease Center. The authors thank Dr. Theodore Kotchen for critical review of this manuscript and Drs. Ellen Blank, Kenneth Lee, Alfonso Martinez, Grzegorz Telega, Farhat Khan, Erin McGuire, and Karen Sherry, RN, for their assistance in recruiting patients.