and follow-up information, including the development of pouchitis after their ileal pouch- anal anstomosis (IPAA). Results: 19 of 22 study patients (male/female ratio: 11/11, mean age: 38 years) had an IPAA and 3 had Brooke ileostomies 21 of 22 control patients (mlae/ lemale ratio: 11/11, mean age: 40 years) had an IPAA and 1 had a Brooke ileostomy. UC patients with fissunng ulcers (range: 1-3 per patient) more frequently had pancolitis (91% vs. 55%, p=0.01), active serositis (41% vs. 5%, p=0,003), and appendiceal involvement (64%vs 27%, p = 0,02) in comparison to control patients without fissuring ulcers. Backwash ileitis was present in 4 of the study and 2 of the control patients (p>0.05). None of the patients in either group developed clinical or pathologic evidence of Crohn's disease (mean fallow-upperiod: 42 months, range: 7-121 months). However, pouchins developed signifi- cantly more often in the study group than the control group (50% vs. 9%, p<0.001). Conclusions: Early superficial fissuring ulcers in patients with fulminant UC do not necessarily imply a diagnosis of Crohn's disease. However, the presence of fissuring-type ulcers in patients with otherwise typical UC denotes a subgroup with more severe, extensive disease and a higher risk of appendiceal involvement and pouchitis following IPAA S1342 MRI Is a Patient Friendly and Accurate Alternative to Ileocolonoscopy in Determining Disease Activity in Crohn's Disease Jasper Florie, Daan W. Hommes, Sander J. H. Van Deventer, Cristina Lavini, Johan S. Lamens, Jaap Stoker Background: Assessment of disease activity is crucial for diagnosis and management of Crohn's disease (CD). Ileocolonoseopy (IS) is the gold standard for appraisal of localization andseventyof mucosa[ damage, but this procedure is invasive and poorly tolerated. Moreover, the terminal ileum is often not accessible due to inflammatory stenosis. To evaluate the use of non-invasive MRI we compared MRI o[ the small bowel and colon without duodenal intubation (MRSC) with IS in determining disease activity in CD. Methods: From February Io November 2002, 25 CD patients undergoing IS for disease severity assessment were included. Dunng IS location, CD endoscopic index of severity (CDEIS) and overall grading of seventy (no, mild, moderate or severe inflammation) were scored. MRSC was performed within 2 weeks before or after IS. Bowel preparation at MRSC consisted of no food ingestion for 4 hours and oral intake of minimal 1 L of water 2 hours prior to scanning. MRSC with a 1 5 T scanner, included coronal and axial Tme-FISP (pre-contrast) series and Tl-weighted senes after iv. Gadolinium. Patient experience at IS and MRSC was determined. All images wereblindly analyzed for location and overall grading (no, mild, moderate or severe inflam- mation). Exact correlation or one level of difference in grade of severity between MRSC and iS were considered appropriate. Results: Of 25 patients, 6 showed severe, 6 moderate, 5 mildand 8 no activity at IS. In all patients image quality was adequate. In 21 patients there was an exact correlation (n = 14) or one level of difference (n = 7) in agreement between MRSC and IS, There was a significant correlation between grading at MRSC and IS (P-<0.05 r = 0.45), although not between MRSC and CDE1S. In 4 patients discrepancies in degree of disease activity were observed: 1) strong enhancement of the (proximal) ileum at MRSC, 2) absence of enhancement at MRSC 3) missing of a fistula at MRSC and 4) a stenosis of the neoterminalileum not seen at IS. In 5 patients intubation of the terminal ileum was impossible at IS. At MRSC in 4 of these patients a stenosis of the terminal deum was seen. The presence of inflammation seen at IS and MRSC correlated in 87 of 117 bowel segments. 5 of 117 segments were found positive at CS but not at MR. 25 of 117 were found positive for inflammation at MR but not at IS All patients, except one, preferred MRSC. Conclusion: MRSC is a feasible, simple, non-invasive technique for disease assessment in CD patients, demonstrating an excellent tolerability and accuracy. $1343 Muhicenter Prospective Validation of a Simple Endoscopic Score for Crohn's Disease (SES-CD) Marco Daperno, Geert D'Haens. Gert Van Assche, Filip Baert, Jean-Frederic Colombel, Philippe Bulois, Vincent Maunoury, Raflaello Sostegni, Rodolfo Rocca, Angelo Pera, Annemie Gevers, Paul Rutgeerts Background: Endoscopic healing has become one of the major goals of Cmhn s disease (CD) therapy. The gold standard for CD endoscopic evaluation is the Crohn s disease endoscopic index of severity (CDEIS), whmh is time consuming and difficult to use. SES- CDhas been proposed as a simpler ahernauve 11]. Aims: To evaluate SES-CD reproducibility and clinical correlations. Methods: Clinical history, CDAI and IBDQ were recorded. Blood was drawn for CRP and blood count. CDEiS and SES-CD were independendy scored by a pair of experienced endoscopists. The following lesions were assessed in the 5 classic iieocolonic segments: Results: Sixty-two patients were studied. The mean number of ileoco- Ionic segments explored per procedure was 4.6 (range 3-5). lnterobserver agreement for SES-CDvariables was good to excellent for all segments, with kappa values ranging between 076 and 100. The interobserver correlation for SES-CD was significantly closer than for the CDEIS (r=0.97 vs 0.87, p<0.0001). Correlations of SES-CD to CRP and CDAI were significant (p<0.002), but not very close (r=0.39 and 0.41), and not srgnificantly better than CDEIS correlations (0.26 and 0.36). Conclusions: This prospective validation study confirmed the reproducibility of the SES-CD. SES-CD is at least as reliable as CDE1S, but is srmpler to be scored. Further validation is ongoing. [1] Daperno M, et al. Gut 2002;51:A56 Score 0 1 2 3 Ulcenz None <0.5 cm 0.5-2 cm >2 cm Ulcmzt~ 0% <10% 10-30% >30% rdrface Affectedsurface 0% <50% 50-75% 9 Nan'owinp None Single,can be passed MullJp~e, can be passed Cannotbe passed SES-CD calculation= (sum of all variables)-1.4x(nuraber of affected segments). S1344 Cytokine/Chemokine Transcript Profiles Reflect Mucosal Inflammation in Crohns Disease and Predict Relapses After Steroid-Induced Remission Andreas Staflmach, Carsten Schmidt, Stefan C. Meuer, Thomas Giese Background and Aims: Cytokines play an essential role in the pathogenesis of inflammatory bowel disease (IBD). Therefore reliable measurement of cytokine transcripts could serve as estimate for inflammatory activity in mucosal biopsies. Methods: Expression of 30 pro- inflammatory gene transcripts were studied in mucosal biopsies from 70 patients with active IBD (Crohns disease (CD), n=45; ulcerative colitis (UC), n=25) and 12 patients with non-specific colitis (ischemic colitis, infectious colitis). 16 patients with non-inflammatory conditions served as controls. In all patients with inflammatory disorders paired samples from inflamed and non-inflamed areas were analyzed. Transcripts were quantified using Real-time PCR. Results: Compared to not-inflamed mucosa the vast majority of active CD tissue samples expressed significantly elevated .transcript levels for IL-I~, IL-8, MRP-14, MIP2or and MMP-1. Moreover, increased cytokine transcript levels were detected in both active UC and specific colitis Importantly, TNF-c~-, 1FN-3*-,1L-23-, and IL-27-transcripts were found increased in active CD only. Transcript levels (IL-1I~, IL-8, MRP-14, MMP-1) correlated with clinical disease activity (CDAI) and endoscopic scoring indices. Medical treatment induced stable remission m 14 of 20 patients which was paralleled by a reduction of increased transcript levels. Six out of six patients without normalization of MIP2tx-, MRP- 14-, TNF-a- and IL-l(3-transcripts developed an early relapse during follow-up. Discussion: Elevation of pro-inflammatory cytokine transcripts in active CD may underly disease reactiva- tion and chronicity. Real-time PCR quantification represents a simple and objective method for grading inflammation of intestinal mucosa and may be useful to identify patients who would benefit from anti-inflammatory remission maintenance. $1345 Faecal Calprotectin as an aid to Diagnosis in Intestinal Inflammation Sunfl Dolwani, Julie Wassefl, Magdalena Metzner, Helen Losty, Audrey Yong, Brian Lawrie, Anthony B. Hawthorne Background/aims: The nentrophil derived protein Calprotectin has previously been found to be raised in intestinal inflammatory conditions. We aimed to evaluate the discriminant value of a stool calprotectin assay in predicting the likelihood of an abnormal result on a Barium follow through examination (BaFT) in patients being investigated for abdominal pain and or diarrhoea Patients and Methods: Patients being investigated for abdominal pain and or diarrhoea and undergoing a BaFT as part of their workup (n=65) provided a one off stool sample for estimation of calprotectin level This was compared with patients with known active Crohn's disease (positive controls), normal health}?volunteers & patients with irritable bowel syndrome (IBS) as negative controls. The biochemist performing the assay was blinded to all clinical details. Other clinical and laboratory indices such as ESR, CRP and CDAl were assessed concomitandy. Results: The Median level of calprotectin in the active Crohn's group (n=23) was 226.5 mg/g of stool compared to a median of 17.3 in the group with iBS (n=27) and 10.9 in normal healthy controls (n=24) A sensitivity of 94%, specificity of 68.7% & negative predictive value of 97% for the stool calprotectin assay. Of the 6 patients with 1BD & a normal barium follow through, 5 had colonic Crohn's disease & 1 had Ulcerative colitis on further investigation. Conclusion: Patients being investigated in a gastroenterology clinic for diarrhoea and or abdominal pain do not need small bowel radiology to rule out Crohn's disease if their stool calprotectin level is <60 mg/g. Patients with normal & abnormal BaFT with Calprotectln levels < or > 60 mcg/g of stool BaFT Normal BaFT Abnormal Calprotectin < 60 mcglg 33 1 CalproCectin 960 mcglg 15 16 Of those 15 with rased calprotestinand a normal BaFT (6 IBD, 4 IBS, 4 others, 1 awallJng colo- noscopy) S1346 Correlation of Short Inflammatory Bowel Disease Questionnaire (Sibdq) with Clinical Status in Routine Management of Crohn's Disease (Cd) Patients Including Ostomates Tan Attila, Kia Saeian, Josh F. Knox, Jeanne Emmons, Lisa Calabrese, Darius Rose, Devang Prajapati, Subra Kugathasan, Michael F. Otterson, David G. Binion Background and Aims: Routine measurement of disease activity in patients with CD is hampered by lack of uniformly accepted or standardized instruments, requirements for diaries, as well as excluding patients with ostomies. We compared the ability of a validated health related quality of life tool (Short Inflammatory Bowel Disease Questionnaire (SIBDQ; Am J Gastroenterol 1996; 91:1571-8), with a validated short disease activity index (Harvey- Bradshaw disease activity index (HB)), in their ability to predict significant clinical outcome, namely hospitalization and surgery in a tertiary referral CD population. We also characterized the ability of the SIBDQ to measure clinical status in CD patients with ostomies. Methods: Over a 9 month period, all CD patients seen in a tertiary referral center had SIBDQ and HB scores determined at the time of initial and follow-up visits. Clinical outcome over the next 6 months was reviewed and correlated with both S1BDQ and HB scores, focusing on severe illness (i.e. hospitalizations and surgeries). All CD patients with ostomies were identi- fied, and SIBDQ scores were obtained in these individuals. Correlation of SIBDQ scores between patients with and without ostomies who did and did not require hospitalizations and surgeries was also analyzed. Results: 99 CD patients provided 128 paired SIBDQ and HB measurements. The correlation (R value) between SIBDQ and HB score was 0.78 (p= <0.01). SIBDQ and HB correlated equally well in predicting severe illness, and was significantly different from patients who did and did not require hospitalization or surgery (SIBDQ 41.6 + 12.9 v 54.4 s 11.2; mean _+ SD; p=<0.003; HB 8,2 _+ 0.8 v 3.5 -+ 0.4 (mean _+ SE); p = <0.0001). Out of 297 total CD patients, there were 36 ostomates, A-199 AGA Abstracts