Crohn disease (CD) is a heterogeneous chronic idiopathic

Transcription

1 ORIGINAL ARTICLE The Epidemiology and Phenotype of Crohn s Disease in the Chinese Population Rupert W.L. Leong, MBBS,* James Y. Lau, MBBS, MD, and Joseph J.Y. Sung, MBBS, MD, PhD Background: Inflammatory bowel disease is uncommon in Southeast Asia but is increasing in incidence. The epidemiology and phenotype of Crohn disease (CD) in the Chinese population is not wellknown. The purpose of this study was to determine the incidence, temporal trend, clinical features, risk factors, extraintestinal manifestations, and the treatment of CD in the Chinese population of Hong Kong. Methods: We performed a single-center study of consecutive definite CD cases based on internationally accepted criteria, with strict exclusion of infective enterocolitis. Results: Eighty Chinese CD patients were recruited, characterized by male gender predominance (male:female ratio 2.5:1), no association with ever smoking (OR 1.02, 95% CI: ), absence of familial clustering (0%), high proportion of upper gastrointestinal tract disease proximal to the terminal ileum (19%), and a low proportion of isolated terminal ileal disease (4%). The mean age at diagnosis was 33 years. Forty-five percent of patients had penetrating disease, 18% stricturing disease, and 37% had nonstricturing, nonpenetrating disease. Twenty-five percent of patients had at least 1 extraintestinal manifestation, and there was a high rate of ankylosing spondylitis (9%). The incidence of CD was 1.0 per 100,000 and has increased by 3 fold during the past decade. The age-adjusted incidence was 3.0 per 100,000 (95% CI: per 100,000). Conclusions: The incidence of CD in the Chinese is increasing. There are some notable epidemiological and phenotypic differences between Chinese CD with Caucasian CD including the lack of familial clustering, male predominance, and higher proportion of upper GIT involvement and lower frequency of isolated terminal ileal disease. Received April 8, 2003; accepted March 29, From the *Department of Gastroenterology, The University of New South Wales, Bankstown-Lidcombe Hospital, Sydney, NSW, Australia; and the Departments of Surgery and Medicine & Therapeutics, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, Hong Kong, China. Dr. Rupert Leong was partially supported by an overseas research fellowship from the University of Western Australia. Reprints: Dr. Rupert W.L. Leong, Senior Lecturer in Medicine, The University of New South Wales, Department of Gastroenterology, Level 3 Bankstown-Lidcombe Hospital, Sydney, NSW 2200, Australia ( rupertleong@hotmail.com). Copyright 2004 by Lippincott Williams & Wilkins Key Words: Asia, Chinese, Crohn disease, epidemiology, inflammatory bowel disease (Inflamm Bowel Dis 2004;10: ) Crohn disease (CD) is a heterogeneous chronic idiopathic inflammatory bowel disease (IBD) that affects people worldwide. However, the incidence rates markedly differ geographically and among different ethnic groups presumably due to genetic and environmental factors. 1 Little is known about IBD in South East Asia. 2 5 However, there is growing evidence that CD is increasing in this region. 6 9 A study of the CD epidemiology in Southeast Asia may help increase the awareness of the extent of this disease in the region, help define the natural history of this disease in the Chinese population, assist with local health policy planning and promotion, and assess theories on environmental and genetic risk factors. Several factors impede the study of IBD in Asia, where gastrointestinal infections, absence of an IBD registry, physician unawareness, and attendance to traditional Chinese herbalists, may make it difficult to determine the epidemiology and characteristics of IBD. 10 This study was conducted at the Gastroenterology ambulatory outpatient clinic of the Prince of Wales Hospital, Hong Kong (HK), which services a welldefined catchment population of 1 million people. In HK, 94% of medical care is performed through the public hospital system, and this provides an opportunity to perform a regionbased study on CD epidemiology and phenotype. 11 PATIENTS AND METHODS Patient Selection Consecutive Chinese patients were prospectively recruited from the gastroenterology ambulatory clinic of the Prince of Wales Hospital. All patients were ethnically southern Chinese (Cantonese), with 85% born in HK and 15% were emigrants from southern mainland China. The population consists of middle-income earners, with a median monthly income slightly higher than the median income for the whole of HK. 12 Patients were longitudinally followed from the time of initial diagnosis (inception cohort). The diagnosis of IBD had to be certain, and was made in accordance with previously established internationally ac- 646 Inflamm Bowel Dis Volume 10, Number 5, September 2004

2 Inflamm Bowel Dis Volume 10, Number 5, September 2004 Chinese Crohn Disease cepted criteria 13 based on clinical, endoscopic, histopathological, and radiologic findings. Typical features of CD included skip lesions, cobblestoning, noncaseating epithelioid granulomas, transmural inflammation, deep ulceration, and stricturing or fistula formation. Diagnosis had to be of at least 6 months duration. Indeterminate colitis, microscopic colitis, infective enterocolitis and Behcet s disease were excluded. Data on the annual incidence of UC over time is presented for comparison. Stool microscopy and culture, bacterial and amoebic serology, acid-fast staining of biopsies and mycobacterial cultures were performed to exclude infectious enterocolitis. Phenotypic classification of patients was performed according to the Vienna Classification, which grouped CD subjects according to age at diagnosis (A1: <40 years; A2: 40 years), location of disease (L1: terminal ileum with or without spillover into the cecum; L2: colon; L3: ileocolon; L4: any involvement proximal to the terminal ileum), and disease behavior (B1: nonstricturing, nonpenetrating; B2: stricturing; B3: penetrating). 14 Intestinal tuberculosis (TB) was excluded in 2 patients with terminal ileitis who failed to respond to anti-tb therapy and subsequently improved on immunosuppressive therapy. They were recruited following a revised diagnosis to CD. The gastroenterology ambulatory clinic at the hospital provided a weekly service dedicated to IBD patients. Due to the incentive of a nominal cost of treatment, 94% of the population requiring health care attended the public hospital system rather than to a private practice. 11 There were no private gastroenterologists in the local catchment district, and Chinese IBD patients perceived a lower level of expertise in the management of IBD among community general practitioners. 15 Also, the HK health system is district-based with funding based upon regional catchments. Therefore patients who did not belong to the region were referred to their district hospital and were excluded from the study. Patients once recruited remained in the study unless the diagnosis was disproved. Modern diagnostic procedures were provided to all patients irrespective of socioeconomic class at no cost. Cases were prospectively entered into the IBD database and a further search was made using ICD9/ICD10 discharge codes of the hospital and from the Endoscopy Unit to ensure case ascertainment. These factors contributed to the best possible and reasonably accurate representation of CD patients in the district served by the hospital. The study design meets the study design criteria as suggested by Farrokhyar et al. 16 The denominator used in the calculation of annual incidence was derived from the Census and Statistics Department of HK. 12 The 2001 population figure for the catchment area (Shatin and Sai Kung districts) was 956,323. For the years inbetween the census when population data were not available, a rate of 1% per annum change in population was calculated and this was the mean annual rate of rise as determined by the Census and Statistics Department during the study period. 12 Age adjustment was performed for the age group years that accounted for 14% of the population to determine an ageadjusted incidence. The year of diagnosis of CD rather than the year of symptom onset was used to determine incidence to avoid recall bias. Ulcerative Colitis Patients and Community Controls Data for ulcerative colitis (UC) was collected to determine the incidence and temporal trend in comparison with CD. This was to exclude reclassification of UC into CD or vice versa as the reason for and increase or decrease in CD incidence. A total of 122 UC subjects were recruited in the same way as for CD from the same population over the same time period. One thousand healthy community controls were recruited from the same district and were obtained from a community health survey (unpublished data). From this control population, matching at a ratio of 3:1 by sex and age (± 3 years) was performed to compare the risk factors in the development of CD. Statistics Data collected from all patients included demographics, smoking history, level of education, occupation, family history of IBD, symptoms, investigations, disease distribution and behavior, extra-intestinal manifestations, complications, and treatment. Statistical methods for comparison of continuous normally distributed data were by the t test, comparison of dichotomous data by the 2 test, and logistic regression to estimate odds ratio of risk factors. Ethical Considerations Signed informed consent was obtained from patients and the study was approved by the Chinese University of Hong Kong ethics committee. RESULTS Demographics Eighty consecutive patients with CD were identified and recruited. The median follow up time was 5.6 years. There were 57 males and 23 females (71% males; 95% CI: ; Table 1). The mean age of diagnosis was 33.1 years with 78% of all patients presenting below the age of 40 years. None of the CD patients had a family history of IBD. There was a suggestion that the distribution of the age of onset may be bimodal, similar to that of Caucasians with CD, 17,18 with the main peak of onset in the third decade, and a secondary peak in the sixth decade (Fig. 1). A higher socioeconomic class had been previously implicated in the development of IBD. 19 In this cohort, 45% of people had received a tertiary (university or technical college) education, but only 24% were employed as a professional or skilled worker Lippincott Williams & Wilkins 647

3 Leong et al Inflamm Bowel Dis Volume 10, Number 5, September 2004 TABLE 1. Demographics of Chinese Crohn Disease Patients Gender: male, female 57, 23 Mean age diagnosis: year (SD) 33.1 (14) Country of birth: (%) China 15 Hong Kong 85 Education (%) Primary or below 15 Secondary 12 Apprentice 12 Tertiary (university or college) 45 Occupation (%) None, homemaker 20 Student 30 Unskilled labor 26 Skilled labor 12 Professional 12 Smoking: (%) Current smoker 13 Ex-smoker 13 Nonsmoker 74 Disease median duration: years 5.6 Presentation (%) Diarrhea 65 Abdominal pain 65 Rectal bleeding 51 Weight loss 45 Fever 17 Anemia 12 Extraintestinal 4 in the CD group. Re-analysis selecting only patients recently diagnosed with CD from 1997 onwards, consisting of a group less likely to be biased from smoking-related deaths and drop out, demonstrated that the risk of developing CD in eversmokers remained low and insignificant (OR 1.47; 95% CI: ; Table 2). These data suggested that, overall; ever smoking was unlikely to be a risk factor in the development of CD in the Chinese. Presentation and Investigations The median time to diagnosis was 9 months, which may have underestimated the incidence of disease in any given year, but was insufficient to account for the large rise in incidence over the decade. Symptoms at presentation were diarrhea (65%), abdominal pain (65%), rectal bleeding (51%), weight loss (45%), fever (17%), anemia (12%), and extraintestinal manifestations (4%). Other presentations included small bowel obstruction, fistula in ano, abdominal mass, and protein-losing enteropathy. All patients had a colonoscopy and 80% also had small bowel enema. Phenotype Penetrating disease occurred in 45% of all patients. Stricturing disease was observed in 18% of patients, and 37% of patients had nonstricturing, nonpenetrating disease. Disease distribution was as follows: 4% ileum, 30% colonic, 44% ileocolonic, and 19% had confirmed involvement proximal to the terminal ileum (Table 3). Of the latter group of patients designated as having upper GIT disease, 7 had small bowel studies disclosing strictures and/or fistulae, and 8 had histo- The annual incidence rates of both CD and UC in Hong Kong have increased over the past decade (Fig. 2). The incidence of CD increased more that three-fold from 0.3 per 100,000 in 1989 to the present rate of 1.0 per 100,000 (95% CI: per 100,000). The rate of rise was greatest in the mid to late 1990s. The incidence for males was 1.3 per 100,000 (95% CI: per 100,000) and for females 0.6 per 100,000 (95% CI: per 100,000). The age-adjusted incidence for the 20- to 29-year-old age group was 3.0 per 100,000 (95% CI: per 100,000). Ever smoking was not a risk factor in the development of CD in this study that included 13% current smokers, 13% ex-smokers, and 74% nonsmokers. The OR of ever-smokers in the development of CD compared with age and sex matched healthy controls was only 1.02 (95% CI: ). However, the CD patients were less likely to be current smokers (OR = 0.46; 95% CI: ) and more likely to be ex-smokers (OR = 7.56; 95% CI: ) compared with controls. Smokers who died of smoking-related illnesses may have contributed to a lower proportion of smokers FIGURE 1. Distribution of the age of diagnosis of Crohn disease in the Chinese population Lippincott Williams & Wilkins

4 Inflamm Bowel Dis Volume 10, Number 5, September 2004 Chinese Crohn Disease FIGURE 2. The temporal trend of the incidence of Chinese Crohn disease (CD, solid line) and ulcerative colitis (UC, dotted line) up to The numbers of CD and UC patients in each year group are tabulated. Patients recruited in 2002 were not included in this figure. pathological findings from endoscopic or laparotomy. Of those with upper GIT disease, 1 patient had disease limited only proximal to the terminal ileum, 6 also had terminal ileal involvement, 3 also had colonic involvement, and 4 had ileocolonic involvement. The presence of upper GIT involvement significantly predicted stricturing CD (P = 0.009) or penetrating disease (P = 0.034). Additionally, 36% of CD patients had documented perianal fistulae, fissures or abscesses. Forty (50%) patients had granulomas on histology. Patients who had granulomas were not found to have tuberculosis (TB) either from polymerase chain reaction, culture, or failure to respond to tuberculosis-specific therapy. Extraintestinal Involvement At least 1 extraintestinal manifestation of IBD was present in 25% of patients. Ankylosing spondylitis was diagnosed by back pain or stiffness along with typical radiologic features and was found in 9% of patients, 6 males and 1 female. HLA- B27 was not routinely tested in these patients. In addition, 9% of patients had monoarticular or pauciarticular arthritis and this prevalence is similar to Western CD cohorts. Eye involvement occurred in 4% (2 patients with anterior uveitis and 1 with episcleritis). Skin involvement occurred in 2% of patients (1 with pyoderma gangrenosum and 1 with erythema nodosum). Gallstones were detected in 19% of CD patients. Abdominal ultrasound was not performed routinely and the high percentage of gallstones was likely to be due to the selection of patients investigated for abdominal pain. Treatment and Outcome The cumulative rate of surgery is illustrated in Figure 3 and demonstrated 29% of patients having had major resection surgery or strictureplasty. The indications were obstructive stricture (11%), perforation or abscess (9%), and hemorrhage (4%). In addition, 4% of patients had defunctioning colostomies for perianal fistulas or distal strictures. Another 19% of patients had perianal surgery, mostly fistulotomy. Twenty-six percent of patients had been on immunomodulators including TABLE 2. Relative Risk (and 95% Confidence Interval) of Smoking in Crohn Disease Patients Compared to Matched Controls Indicating the Lack of Association Between Ever-Smoking and CD Ever-Smoker Current Smoker Ex-Smoker Never-Smoker CD 1.02 ( ) 0.46 ( ) 7.56 ( ) 0.97 ( ) CD Recent 1.47 ( ) 0.70 ( ) 8.14 ( ) 0.68 ( ) *A higher proportion of CD patients quit smoking compared to controls and shifted from current smokers to ex-smokers. Re-analysis of recent CD cases diagnosed after 1997 to reduce bias of smokers dropping out due to smokingrelated illnesses Lippincott Williams & Wilkins 649

5 Leong et al Inflamm Bowel Dis Volume 10, Number 5, September 2004 TABLE 3. Disease Phenotype According to the Vienna Classification Age at diagnosis (%) A1: <40 years 79 A2: 40 years 21 Disease location (%) L1: terminal ileum 4 L2: colon 30 L3: ileocolon 44 L4: upper GIT 19 Unclassified 3 Disease behavior (%) B1: nonstricturing, nonpenetrating 37 B2: stricturing 18 B3: penetrating 45 azathioprine, 6-mercaptopurine, and methotrexate. Only 2 patients have had anti-tnf therapy to date due to the high cost and limited availability of biologic agents in HK. None of the patients have had colorectal malignancies. There were 4 deaths during the follow up period, none directly attributed to CD. DISCUSSION This was a descriptive epidemiological study on CD that focused on a homogenous Chinese population. Of note was the increase in CD incidence, with an age-adjusted rate of 3.0 per 100,000. The increasing annual incidence of CD has also occurred elsewhere. 20,21 Reasons for this increase may have been from improved physician awareness of CD, better availability of diagnostic modalities, reclassification of disease, or a true rise in disease incidence. However, the Division of Gastroenterology and the Endoscopy Unit were established a decade before the rapid rise in disease incidence. The diagnostic facilities, staff, equipment and diagnostic criteria have remained unchanged during this time. There was also a similar rise in the incidence of UC during the same time period suggesting an overall increase in IBD rather than reclassification of UC into CD. These facts therefore supported a true rise in the incidence of CD. This was an ambulatory clinic-based study and there was a possibility of an underestimation of the incidence caused by milder cases presenting to alternative non-western health practitioners. However, another study showed that the overall proportion of Chinese IBD patients using alternative therapies to be no different to that of Caucasian IBD patients. 15 Also, the structure of the HK health system was that even patients with minor gastrointestinal complaints often presented to tertiary hospitals or were referred to the hospital outpatient clinics, due to an absence of a well-defined public sector of general practitioners and the resulting cost saving. So patients of all degrees of CD severity were recruited. Several key differences were noted in the Chinese CD patients. Smoking is the best known and consistent environmental risk factor in the development of CD. In the Chinese population, ever smoking was not a risk factor for CD with an OR of 1.02, compared with an OR of in Western populations Another characteristic was the male gender predominance. This is the reverse of the gender distribution in Western countries that tend to have a female-to-male ratio above 1. 15,25 Male gender predominance in CD is also noted in some countries with a low CD incidence. 10,26 28 Another risk factor for CD is a positive family history, which may in part be due to genetic polymorphisms of the NOD2/CARD15 gene. 29,30 Our center previously reported that the NOD2/CARD15 gene variants to be absent in the Chinese CD population. 31 Not surprisingly, there was no familial clustering of IBD in this cohort. The location of CD in the GIT in the Chinese differed to Caucasians, with a higher proportion of Chinese having upper gastrointestinal tract (GIT) disease (19%) compared with only 0.5% to 13% in Caucasians. 32 Most of the Chinese patients with upper GIT CD had presentations of proximal small bowel obstruction and perforations, rather than incidental or minor radiologic abnormalities, suggesting that small bowel CD complications occurred more frequently in the Chinese. The location of disease according to the Vienna classification remains stable on long-term follow up studies, 33 indicating that upper GIT involvement to be a genuine phenomenon in the Chinese CD population. Conversely, fewer Chinese CD patients had isolated terminal ileal disease only 4% compared with 30% in a Western series. 34 Other notable findings in Chinese CD include a high proportion of penetrating (45%) and stricturing complications (18%) and 9% of patients with ankylosing spondylitis. Chinese patients differed to Koreans in whom spondylitis was rare, affecting only 1 of 52 of Korean CD patients. 35 FIGURE 3. The cumulative rate of major surgery for Crohn disease Lippincott Williams & Wilkins

6 Inflamm Bowel Dis Volume 10, Number 5, September 2004 Chinese Crohn Disease The difficulties in studying IBD epidemiology are well recognized, and affect studies of all populations. 16 For this study to be robust, the diagnosis of CD had to be definite, the cohort was representative of the catchment population, there was high referral rate of IBD to the ambulatory clinic, there was high case ascertainment, and the median follow-up time was adequate. In conclusion, CD is an emerging disease in the Chinese community with a 3-fold increase in incidence during the past decade. It affects predominantly the young male population, and smoking is not a risk factor. Phenotypic differences include more upper GIT involvement, less isolated terminal ileal disease, and higher frequency of ankylosing spondylitis. The lack of familial clustering makes a genetic basis for this disease less likely in the Chinese. The reasons for the increasing incidence of CD in HK remain unclear. Possible mechanisms for this increase, such as Westernization of the Chinese diet and culture, changes in intestinal infections, and flora and host immune interactions, are important areas that need to be studied. ACKNOWLEDGMENT Control cohort data were provided by Ms. Jessica Ching and Professor Francis Chan, Department of Medicine and Therapeutics, the Chinese University of Hong Kong. REFERENCES 1. Moum B, Ekbom A. Epidemiology of inflammatory bowel disease methodological considerations. Dig Liver Dis. 2002;34: Sung JJ, Hsu RK, Chan FK, et al. Crohn s disease in the Chinese population. An experience from Hong Kong. Dis Colon Rectum. 1994;37: Lee YT, Sung JJ, Poon P, et al. Association of HLA class-ii genes and anti-neutrophil cytoplasmic antibodies in Chinese patients with inflammatory bowel disease. Scand J Gastroenterol. 1998;33: Lau PW, Boey J, Lorentz TG. Ileo-anal pouch procedure: experience in the Chinese population. AustNZJSurg. 1991;61: Sung JY, Chan KL, Hsu R, et al. Ulcerative colitis and antineutrophil cytoplasmic antibodies in Hong Kong Chinese. Am J Gastroenterol. 1993;88: Tan CC, Kang JY, Guan R, et al. Inflammatory bowel disease: an uncommon problem in Singapore. J Gastroenterol Hepatol. 1992;7: Higashi A, Watanabe Y, Ozasa K, et al. Prevalence and mortality of ulcerative colitis and Crohn s disease in Japan. Gastroenterol Jpn. 1988;23: Morita N, Toki S, Hirohashi T, et al. Incidence and prevalence of inflammatory bowel disease in Japan: nationwide epidemiological survey during the year J Gastroenterol. 1995;30(suppl 8): Law NM, Lim CC, Chong R, et al. Crohn s disease in the Singapore Chinese population. J Clin Gastroenterol. 1998;26: Yang SK, Loftus EV Jr, Sandborn WJ. Epidemiology of inflammatory bowel disease in Asia. Inflamm Bowel Dis. 2001;7: Hospital Authority of Hong Kong. Overview of healthcare services in Hong Kong. Published on the Hospital Authority Web site, Available at: ha%5fview%5ftemplate%26group%3dhci%26area%3dhko (accessed 7 March 2003). 12. Census and Statistics Department of Hong Kong. Population and Vital Events. Hong Kong in Figures, Available at: hk/censtatd/eng/hkstat/fas/pop/pop_vital_index.html (accessed 7 March 2003). 13. Lennard-Jones JE. Classification of inflammatory bowel disease. Scand J Gastroenterol Suppl. 1989;24: Gasche C, Scholmerich J, Brynskov J, et al. A simple classification of Crohn s disease: report of the Working Party for the World Congresses of Gastroenterology, Vienna Inflamm Bowel Dis. 2000;6: Leong RWL, Lawrance IC, Ching JYL, et al. Knowledge, quality of life and use of complementary and alternative medicine and therapies in inflammatory bowel disease: a comparison of Chinese and Caucasian patients. Digest Dis Sci. 2004;49(accepted). 16. Farrokhyar F, Swarbrick ET, Irvine EJ. A critical review of epidemiological studies in inflammatory bowel disease. Scand J Gastroenterol. 2001; 36: Rose JD, Roberst GM, Williams G, et al. Cardiff Crohn s jubilee. The incidence over 50 years. Gut. 1988;29: Haug K, Schrumpf E, Halvorsen JF, et al. Skjolingstad and the Study Group of Inflammatory Bowel Disease in Western Norway. Epidemiology of Crohn s disease in Western Norway. Scand J Gastroenterol. 1989; 24: Sonnenberg A. Occupational distribution of inflammatory bowel disease among German employees. Gut. 1990;31: Bergman L, Krause U. The incidence of Crohn s disease in Central Sweden. Scand J Gastroenterol. 1975;10: Kyle J. Crohn s disease in the North-eastern and Northern Isles of Scotland: an epidemiological review. Gastroenterol. 1992;103: Shivananda S, Lennard-Jones J, Logan R, et al. Incidence of inflammatory bowel disease across Europe: is there a difference between north and south? Results of the European collaboration study on inflammatory bowel disease (EC-IBD). Gut. 1996;39: Lindberg E, Tysk C, Andersson K, et al. Smoking and inflammatory bowel disease. A case control study. Gut. 1988;29: Franceschi S, Panza E, La Vecchia C, et al. Nonspecific inflammatory bowel disease and smoking. Am J Epidemiol. 1987;125: Lee FI. Costello FT. Crohn s disease in Blackpool incidence and prevalence Gut. 1985;26: Yoshida Y, Murata Y. Inflammatory bowel disease in Japan: Studies of epidermiology and etiopathogenesis. Med Clin North Am. 1990;71: Triantafillidis JK, Emmanouilidis A, Manousos O, et al. Clinical patterns of Crohn s disease in Greece: a follow-up study of 155 cases. Digestion. 2000;61: Yao T, Matsui T, Hiwatashi N. Crohn s disease in Japan: diagnostic criteria and epidemiology. Dis Colon Rectum. 2000;43(suppl 10):S85 S Ogura Y, Bonen DK, Inohara N, et al. A frameshift mutation in NOD2 associated with susceptibility to Crohn s disease. Nature. 2001;411: Hugot JP, Chamaillard M, Zouali H, et al. Association of NOD2 leucinerich repeat variants with susceptibility to Crohn s disease. Nature. 2001; 411: Leong RW, Armuzzi A, Ahmad T, et al. NOD2/CARD15 gene polymorphisms and Crohn s disease in the Chinese population. Aliment Pharm Ther. 2003;17: ) 32. Lammers C Crohn s disease of the upper gastrointestinal tract. In: Allan RN, Rhodes JM, Hanaruer SB, et al (eds). Inflammatory Bowel Disease. London: Churchill Livingston, London, pp Louis E, Collard A, Oger AF, et al. Behaviour of Crohn s disease according to the Vienna classification: changing pattern over the course of the disease. Gut. 2001;49: Lowes JR Crohn s disease of the large intestine. In: Allan RN, Rhodes JM, Hanaruer SB, et al (eds). Inflammatory Bowel Disease. London: Churchill Livingston, pp Suh CH, Lee CH, Lee J, et al. Arthritic manifestations of inflammatory bowel disease. J Korean Med Sci. 1998;13: Lippincott Williams & Wilkins 651