MlCl 125 THE FLORENCE INFLAMMATORY BOWEL DISEASE (TBD) STUDY AN UPDATE OF MORTALITY FOLLOW-UP (1978-1999) G. MasaIal, G. TralloriZ, M. Cemtil, S. Bagzolil, C. Saiwal, F. PaciruZ,G. d’Albasm2, 3 andD. Pallil lEpidemiology Umt, CSPO,Florence; 2Gas~entemlogy DepaRment, CareggiHospital, Florence; 3Regional IBD Referral Center, Florence. CSPO Backgmundand aim A divergent panemof mortality for smoking relatedcauses in ulcerative colitis fUCl and Crohn’s disease(CD) oatientsanda sliehtlv increased colorectal cancer mortalitv were tde &in results of a previ&sly &ported mc&ality&iy on a cohort of 920 IBD patient 1 residing in Florencefollowed till December 1996(Palli et al. GUT 1998;42:175-179). An update of mortality follow up at December 1999was carried out tare-evaluate theseresults. Material and methods:Overall 689 UC and 231 CD patientswere traced in thelocal town offices andin the Regional Mortality Registry. Six patients(0.6%) were lost to follow-up. We observed 104 deaths, that were comparedwith expecteddeathsestimated on the basis of sex-, age- andperiod- national mortality rates applied to the individual personyears at risk. The period at risk to estimateexpected deathsdueto colorectal cancer was censored at tbedate of surgery for tbe 52 (5.7%) subjectswho underwenttoti c&ctomy and/or rectum resection.Standardised mortality ratios (SMRs) (observed/expected deaths)and 95% confidenceinterwds (CI) were computed for overall andcanse specific mortality. Results: In UC patients overall mortality (SMR 0.72; 95% CI 0.6-0.9) andlug cancer modity (SMR 0.26,95% CI 0.03-0.9)were reducedwhile in CD patientp generalmortahty (SMR 1.70;95% CI 1.2-2.4). cancer mortality (OR 2.20; 95% CI 1.2-3.6)and in particular lung cancer mortality (SMR 3.42; 95% CI 1.1-8.0)were increased A modestand not significantly increasedcolorectal cancer moality emerged in UC patientsbased on 4 deaths(OR 1.52;95% CI 0.4-3.9). Only a deathdueto colon cancer was identifiedamongCD patient?. An increased mortality directly relatedto tbe disease was evident in CD patients(OR 5.26; 95% CI 2.1-10.8for the group of gastrointestinal disease causes). Conclusions:The divergent patternof mortality possibly related to differences in smoking habitsbetweenCD andUC patientswas contirmed. nlhe risk of dying ti’om colorectal cancer in tbis populationbasedstudy appearsto be only modestly increased. 126 BOWEL ULTRASOUND (US) IN ASSESSMENT OF CROHN’S DISEASE (CD) AND DETECTION OF RELATED SMALL BOWEL STRKTURBS. A PROSPECTIVE COMPARATIVE STUDY VERSUS X-RAYS AND INTRAOPBRATIVE FINDINGS. F.Parente, G.Maconi, S.Bollani, A.Andnloni, G.Sampietml,M.Cristaldil, N.FmnceschellQ, R.BiancoZ, A.M. Tascbieril, G.Biancbi P-arm. Departments of Gastmentemlagy,IGeneral Surgery, andZRadiolagy,LSaeco University Hospital Milan, Italy. Luigi Saxa Background&aim.Despitethe fact thatbowel ultmsaund(US ) hasrecently proved to beuseful in the assessment of bowel disease. wertaintv persists as to its diaanosficrole in patientswith complicated Cmbn’s disease (CO). We bavk,‘tberefore,pmspecti~elyinvestigatedthe accuracy of US 88compared to X-rays andsmxical findings in assessing disease extent andlocation and detectinzdD-assaciatedsmall b&l strictures. Methods. A seriesof 296 consecutive patients with proven CD admitted to L.Sacco University Hospital between1997and 1999,hating underzone completeradiow&ic evaluation (including small-bowelX-ray, colonoscqy or double con&t bait&enema), were em’olled in tbc study. Bowel US were pzformed in every patient by two experiewed operatorsunaware of tbe results of other diagnosticpmctdurcs. The accuracy of US in detectinn strictures. as conma& with X-my studies,was determinedseparately in two different gm@s of pat&s: 211 p&ents treatedc&servatively (non-operativeCD) and85 patients candidates for surgery for unrespasiveness to medical therapyor complications (operative-CD). Results.Ovemllse&i&y and &e&city of US in assessing the anatomical distribution of CD were 93% and97%, respectively. Tbe extent of ileal disease at US correlated well with that determined by X-rays (I-0.52. p<O.o01)in medically-heated patientsandwith that measured intraoocrativelv in sumicd mtimts +&54.~4).001). Sensitivity, specificity and pasitive predictive valuebf bwciUS in the d&&on df stri&es were 79%, 98% and95% & eon-operativeCD patientsand9(p/o,100%and 100% in operative-CD cases,respectively. Conclusions.In kxtienced bands.US is an accuratete&ioue for assessinn CD extent andlocation andis very helpful in detecting&I bowel shictwes, &ecially in v”y severecases,candidate to surgery. Therefore, its use is justified as primary investigationin CD patientsin whom complicationsare 127 Un unconventionalcase of esophageal Cmhds disease responsive to infllximab therapy GP.Mamne, F.Morace. A.Mamne, V.De Gimlamo. R.FQmpo GastmentemlogyUmt, Ascalesi Hospital of Napoli.Italy Cardinale Ascalesi Cmhn’s disease (CD) is rarely confined to the esophagus We present a case of CD wtb Isolated esophageal localization at clinical presentation. GC, male, 20 years old, referred at March 1999,to our deparbnent complaininaof we&t lass, odvnopbaaia andpimsis refractor to proton pump inbibi&s (PPpO the&y. Hiwas b&pitabzed: l&&t& tests&e all negat& except ESR (86 mm Istbr), CRP (6,s mg/dl) andp-ANCA (positive). Aphtoid ulcer and erosionswith intervening normal mucos~were seenin middle es&arms on uooer endoscoov.Histoloeic manifestationsin biopsy specimensfrom esophagus weresu&cted f&CD. Esoph&s e&&doscopy showed irregular echo pattern in all wall layers. Small towel enema, pancoloscopy andhistologic finding in biopsy were all negative for CD. Infective, iatmgenic,toxic ethiology of oesopbagitis were excluded.We concludedfor a diagnosisof confinedesophageal CD and therapy with prednisone(I mgikgldie’iv), mesalazine (4 g/die) andPPI (20 mgidie) was started.The &senseclinically remitted. Three weeks later,laboratory tests normalizedand esophageal ulcer healed. Hence, maintenance therapy with mesaluine slow release(4 g/die) and PPI (20 mg&e) was stated. In November 2000 GC manifested mucosangubmlent diarrhea,abdominal pain andodynophagy. Upper endoswpy revealedtypical CD lesions of esophagus. Pancoloscopywas positive for ilencecalvalve lesionsand small bowel enema showed terminalileal involvment. Histolwic manifestationsin bioosv specimensfmm colon and esophagus was suggestive for CD. plednisoneat high dose (66 m&/die&v.) was effective for inducing remission but it WBS not at low doses (less than 20 &die iv) for maintaining remission. Immunosuppressive @atmeat (azatbioprine2 mg/kg/die) was started as its steroid-sparingprop&es but it achievedonly poor clinical remission andno endoscopiclesions healing.As the patientwas not respqnsive to conventionaltherapy, he was treatedwith anti TNF-a antibody (Intliximab, REMICADE 0) at doseof Smgkg. After a single infusion of infliximab, no impmvementswere obwwd in ileal endoscopicdisease activity and in oeso~lweal lesionsbut the la& remittedafler repeated dosesof i&ix&b with complete~recovery of &$hageal ulcers sfler the third infusion. Thesepreliminary data show the efficacy of infliximab in the treatmentof 128 APPENDICECTOMY DELAYS THE ONSET OF ULCERATIVE COLITIS, A RETROSPECTIVE STUDY A. D’Arimm, F. Mangum, M. Sanges, R. Be-to, l S. Ga+lo, *P. N&m, G. Vicinanza,T SlkIlO, *A. Rend+ G. Mazzacca. Gastmentemlogyand *GeneralSurgery Units, Faculty of Medicine, Federico U University of Naples,Italy Pobclinicn Universitaio Fedenco II Background and aim The fact that periappewii~eal inflammation is often observed m paheatswith proctitis or l&-sided ulcerative colitis (UC) supports the idea that the appendixmay be closely relatedto tbe pathogenesisof UC. Moreover, appendicectomy seemsto protect againsttbe development of UC andreduces its recurrence. The aim of our retrospective study was to evaluate If appendicectamy may delay tbeonset of UC. Patients and methods:Of the 664 UC outpatients com~utivelv admittedto our Unit. 36 claimed ab&tow of anendtcectomy. Nine of themwere excludedf&n tbe study because they badundergone&endi&tomy less than 12 months from tbe omet of UC. Comcuuently, 27 patients(MIF=9/18) were inchtdedin tbe study. From UC patients without appmdiccctimy we r&domly chose, with theaid of a specific software, 39 subjects as cantml group. The patients’ agesat the onsetof UC andat the appendtcectomywere recorded. Fin&. oatients with aooendicectomv were divided into 2 gmups: with appendiceztomy in _.. childhood(gmup A, n=b; meanage <.3+2.6ym) and in &it a& (group ti; n=l8, meanage 20.9+6.8ym). Results: No differenceof sex disbibuhon was observed in patientswith or without appendiccctomy. The mean time between appadicectomy andthe onsetof UC was 17.3t9.7 yrs. range3-39 yrs. In subjects without appdicsctomy the meanageat the onset of UC was 28.8i8.7 yrs, while in those with appcndicectomy 34.3a10.9yrs @=O.OZ). In gm”p A tbe mea” ageat tbe onsetofUC was 25.1ti.7 yrs, while in group B it was 38.p+9.7yrs @=O.OOl). Conclusions: In patientswith appadicatomy the onsetof UC occurs later thanin thosewithout appendicectomy. This differencewas not obserwd if appendicectomy was paformed in the childhood. Consequently, appendicectomy seemsto delay theonset ofUC only ifperformed in the adult life. A70