A878 AGA ABSTRACTS GASTROENTEROLOGY, VoI. IO8, No. 4 • MESALAMINE(M) IN CROHN'S DISEASE (CD) PATIENTS WITH PREDNISOLONE-INDUCED REMISSION : EFFECT ON STEROID WEANING AND REMISSION MAINTENANCE IN THE YEAR AFTER WEANING : A PLACEBO-CONTROLLED, MULTICENTRE DOUBLE-BLIND TRIAL. R. Modioliani. J.F. Co[ombel,J.L. Dupas, M. Dapoigny. Y. Bouhnik, M. Veyrac, G. Cadiot, B. Duclos, J.C. Soul~, J.P. Gendre, O. Danne, J.Y. Mary on behalf of the GEl'AID, France Corticodependency and early relapse after steroid cessation are major evolutive risks in patients having just undergone remission of CD on pred- nisoloue (Pred). The aim of the present trial was to test if, in such patients, M started at the onset of steroid tapering can 1) increase the percentage of weaning from Fred ; 2) reduce the relapse rate over the year after Pred ces- sation. Methods. 150 patients with a CD attack [Crohn's Disease Activity Index (CDAI) > 200] received oral Pred (1 mg/kg/d) x 3 to 7 weeks; 129 went into clinical remission and were randomized between oral M (Pentasa® '4 g/d, n = 65) or placebo (n = 64), taken till weaning and for 1 year thereaf- ter or until relapse. Results. Groups were similar for clinical and biological items collected initially. 58 and 74 % of patients on placebo and M were weaned from Pred respectively (P = 0.054). At the end of the trial, 9/36 patients on placebo and 14/48 on M were in remission. Kaplan-Meier estimates of time to relap- se after weaning are shown in the figure. % 100 ~. __ -- . Mesalsmine 80 ~-, ~ Pt~Icsbo '1 60 ~ ~ 40 ~ ~ T ~-'~-~ 20 months 0 t i I i ' "l 3 6 9 12 15 Conclusion. M (4 g/d) given to CD patients with a Pred-induced remis- sion when steroids begin to be tapered, and maintained 1 year after wea- ning, tends to increase the frequency of successfull weaning, but does not seem change the rate of subsequent relapse. POUCH-VAGINAL HSTULA: A RARE COMPLICATION AFTER A RESTORATIVE PROCTOCOLECTOMY AND POUCH ILEAL- ANASTOMOSIS M. Molinari, R. Hurst, and F. Michelassi, Chicago, IL. Pouch-vaginal fistulae are rare complications after a restorative proctocolectomy and pouch-ileal-anastomosis. The aim of this study was to review our personal experience with this compfication in order to clarify its incidence and timing, and to review the results obtained with surgical treatment. Between 1/87 and 2/94, the senior author has performed 100 consecutive restorative proetocolectomies with J-pouch ileo-anal anastomosis for ulcerative colitis in 58 male and 42 female patients. Among the 42 female patients (age 14-46 years, mean age 31.5 years), three (7%) developed a pouch.vaginal fistula between one and 56 months later (mean 25 months). Two of these patients also experienced a perineal fistula-in-ano: this preceded the occurrence of the pouch- vaginal fistula in one case and was simultaneous to the pouch-vaginal fistula in the other case. None of these three patients had septic complications at the time of the restorative proetocoleetomy; two of them underwent a full mucosectomy with a protecting ileostomy, which was closed two and one half months later', the last patient underwent a stapled lies-anal anastomosis without protecting ileostomy. An additional patient was referred to our attention 42 months after the original procedure when she developed a complete small bowel obstruction due to a stenotic lies-anal anastomosis. Upon relieving the obstruction by way of mechanical dilatation of the lies-anal anastomosis under general anesthesia, a pouch-vaginal fistula was discovered and diagnosed, These four patients have all undergone attempts at repairing their fistula. Three patients have undergone surgical repair of the fistula by way of a transanal pouch advancement. In two cases this was done without and in one case with the protection of a temporary diverting ileostomy. One of these patients has since had a recurrence of the fistula and she is waiting for another transanal pouch advancement. In the remaining patient, placement of a diverting loop lleostomy for seven months was sufficient for the fistula to heal. The fistula has not recurred in the following three years. • A METHOD OF DETECTING M. PARATUBERCULOSIS USING PCRWITHOUT SOUTHERN BLOT. Erik Mondrow, MD, St Joseph Hospital, Denver, Colorado, J. O'Brien, BS, S. Lewey, DO, A. Guderrez, PhD, P.R. McNally, DO. Fitzsimons Army Medical Center, Aurora, CO. INTRODUCTION: Mycobacterium paratuberculosis (NIP) is the causative agent of Johne's Disease, a chronic enteritis in animals. Hence, MP has been proposed to play an etiologic role in Crohn's Disease, Detection of MP in intestinal tissues is difficult, requiring polymerase chain reaction (PCR) to detect IS900, a DNA insertion element unique to NIP. Prior studies using PCR have relied on Southern Blot Hybridization (SBH) to detect IS900. SBH requires expensive materials, equipment and radioactive markers. We describe a new, more sensitive method of PCR with "nested" DNA primers able to detect MP without SBH. MET.J=IO_D~Oligonucleotide primers Pg0 (5'-GTT CGG GGG CGT C~C TTA GG-3') which ~nplify a 400 bp region of IS900, were used as external primers. The internal primers, 1'90-75 (5'-TCA TGT GGT TGC TGT GTT GG-3'), and P91-811 ('5-GCT CGT CGT CGT TAA TAA CC-3'), were computer designed to amplify the PCR product region Pg0 and Pgl. 5 ul ofa 10~ dilution of MP suspension (ATCC #43544) or 5 ul water (control) was used to "spike" normal human colonic mucosal specimens. Samples digested with proteinase K/ SDS mixture and DNA extracted with phenol and precipitated with ethanol. PCR was performed on samples using the external primers and then internal primers. CONCI.USIONS: This new PCR method clearly identified MP without the need for SBH. • EXPRESSION OF DIFFERENT ISOFORMS OF INTERLEUKIN- 1 RECEPTOR ANTAGONIST (IL-lra) IN NORMAL AND LPS-TREATED RABBITS. L. Monsacchi and F. Cominelli. Department of Medicine, USC-School of Medicine, Los Angeles, CA. We have previously shown that IL-1 and IL-lra are regulated differently in vivo and IL-lra is constitutively expressed in normal tissues including small and large intestine (J. Biol. Chem. 269:6962-6971, 1994). Two isoforms of IL-lra have been described, a secreted (s) form, expressed primarily in monocytes and an intracellular (ic) form expressed in epithelial cells. Little is known about the relative expression of the two isoforms in the intestine. We investigated the expression of slL-lra and icIL-lra in intestinal tissues of rabbits following severe endotoxemia. Colon and small intestine as well as heart, kidney, liver, lung, skin, spleen, and stomach tissues were obtained 4 hrs after i.v. LPS (1.5 mg/kg, n=7) or placebo (1 ml saline, n=5) administration. Two ltg of total RNA were reverse transcribed followed by a polymerase chain reaction using specific primers for slL-lra and icIL-lra. PCR products were visualized on a 3% Nusieve/agarose gel stained with ethidium bromide. Preliminary results suggest that both isoforms are expressed in normal tissues. After LPS administration icIL-lra appeared to be induced in all tissues and in particular in the colon and small intestine. These results suggest that induction of icIL-lra expression during intestinal inflammation may play an important role in regulating immune and inflammatory responses in the gut. A quantitative PCR for rabbit slL-lra and iclL-lra is under development in our laboratory at the present time.