The use of reproducible animal model of intestinal inflammation that can be pharmacologically manipulated represents the important part of ongoing studies of IBD. We documented the inflammatory lesions of the colon in BALB-c mice, using the enema of 0.2% solution of 2,4dinitrofluorobenzene (DNFB) in 4:1 acetone to olive oil solution and combined with previous skin sensitisation. In this murine model, we also demonstrated the effect of methy!prednisolone (1 mg/kg b.w., i.p., daily) on acute inflammatory lesions in significant reduction of mononuelear infiltration within colonic mucosa; compared to control group treated with phosphate buffered saline, in the same maimer (p<0.0I). In similarly designed experiment, eyclosporine A (10 and 20 mg/kg b.w., i.p., daily) has shown to be effective in reducing the observed parameters of inflammation (histopathological score ranging 0-30) (p<0.05). in a dose dependent manner. The results indicate that this murine model of inflammation represents a suitable preparation in studying new therapeutic modalities fo r inflammatory bowel disease. • CHARACTERISTICS OF COLORECTAL CANCER IN PATIENTS WITH INFLAMMATORY BOWEL DISEASE. Pradeep Bansal, Amnon Sonnenberg. Division of Gastroenterology, VA Medical Center and The Medical College of Wisconsin, Milwaukee, WI. Background: It is unknown whether colorectal cancer (CRC) in patients with inflammatory bowel disease (IBD) behaves differently from regular CRC in patients without IBD. To assess the colonic distribution, mortality and risk factors of CRC in IBD, a case control study was conducted comparing CRC in patients with and without underlying IBD. Methods: The Department of Veterans Affairs ('CA) maintains a computerized file of all hospital discharges among US military veterans since 1970. This file accrues the data of one million hospital discharges per year. All patients with IBD and all patients with CRC who had been discharged from a VA hospital between 1981 and 1993 were selected. Individual veterans were identified by their unique social security number and traced in the annual files from 1970 until their last discharge. All previous primary and secondary discharge diagnoses were listed by their 5-digit ICD-code. The influence of various risk factors on the occurrence of CRC in IBD and mortality from CRC in patients with and without IBD was tested by logistic regression analyses. Results: Of the 11,446 sub ects with IBD, 371 had colon cancer (64% with ulcerative colitis and 36% with Crohn's disease). CRC was d agnosed in 52,243 subjects without IBD. The distribution of CRC was not significantly different in the subjects with ulcerative colitis and those without IBD (z2=5.87, df=5, p=0.32), but in Crohn's disease, more cancers were located in the proximal colon (z2=18.10, df=5, p=0.003). The occurrence of CRC in IBD was influenced by the following risk factors: age (odds ratio OR=1.45, 95% confidence interval 1.35-1.57), sclerosing cholangitis (OR=3.41, 2.03-5.73), and history of aspirin consumption (OR=0.84, 0.65-1.09). Sex, race, and type of IBD did not exert a significant influence on the development of cancer. Mortality from CRC was influenced by the following risk factors: age (OR=1.16, 1.14-1.18), male sex (OR=1.24, 1.07-1.45), white race (OR=0.97, 0.96-0.99), and history of aspirin consumption (OR=0.68, 0.65-0.72). Presence of IBD was not associated with a significant influence on CRC mortality (OR=I.00, 0.70-1.43). Conclusions: There is no difference in the characteristics or mortality from colorectal cancer in patients with or without IBD except for a more proximal localization of the CRC in Crohn's disease. Aspirin exerts a protective influence against CRC in patients with IBD similarly as in patients without IBD. Sclerosing cholangitis is associated with a strong risk of developing colon cancer in patients with IBD.