987

Is Bile Reflux-Induced Esophageal Dysmotility Mediated by MucosalStimulation? An Experimental StudyMarcelo S. Rocha, Fernando A. Herbella

Background and Aims: Esophageal motor abnormalities are frequently found in patientswith gastroesophageal reflux disease. The role of bile in reflux-induced dysmotility is stillelusive. Furthermore, it is questionable weather mucosal or muscular stimulation leads tomotor modification. The aims of this study were: (a) analyze the effect of bile infusion inthe amplitude of esophageal contractions and (b) analyze the effect of mucosal vs muscularstimulation. Methods: 18 guinea-pig esophagi were isolated and its contractility assessedwith force transducers. Three groups were studied. In group A (n= 6) the entire esophaguswas used and incubated in 100 mML ursodexocicolic acid for 2 hours. In group B (n=6)the mucosal layer was removed and themuscular layer incubated in 100mML ursodexocicolicacid for 1 hours. In group C (n=6) (control group) the entire esophagus was used andincubated in saline solution. In all groups, five sequential contractions spaced by 1 minutewere measured before and after incubation. Contractions were recorded after KCl 40 mMstimulation. Results: Contractions before incubation did differed among groups (p= 0,006)and averaged 1,319(A),0,306(B) and 1,795(C). After incubation amplitude of contractionwas 0,709 , 0,278 and 1,353 for groups A, B and C respectively. Before incubation thereis no diferrences between groups A and C (p=0,633) there is difference between groups Aand B (p=0,039) and B and C (p=0,048). After incubation there is no differences betweengroups A and B (p=0,134) there is difference between groups A and C (p=0,022) and Band C (p=0,000).When we compare average within groups (before and after) there isdifference only in group A (p=0,030). Conclusion: Our results show that bile exposure mayinduce ineffective esophageal motility and the mucosa seems to take an important role inesophageal motility.

988

The Effects of Cannabinoids on the CaCo-2 Cell Culture Model of IntestinalPermeabilityAbdussalam Alhamoruni, Andrew Lee, Jon Lund, Richard I. Hall, Mike Larvin, SaoirseO'Sullivan

The endocannabinoid system is expressed in the gastrointestinal system, and activation ofcannabinoid receptors in the gut decrease emesis, gastric acid secretion and intestinal motility.However, as yet, the effects of cannabinoids on intestinal permeability have not been estab-lished. The aim of the present study was to examine the effects of cannabinoids on intestinalpermeability using the Caco-2 cell model. Caco-2 cells were grown until fully confluent oninserts in 12-well plates. Transepithelial electrical resistance (TEER) measurements weremeasured as an index of permeability. 50 μM EDTA was applied to inserts to cause a fallin TEER (an increase in permeability) and TEER was measured for the following 4 h (untilTEER had recovered to baseline). The effects of cannabinoids on TEER in combination withEDTA, or alone, were assessed. Potential target sites of action were investigated using thefollowing (all 1 μM); AM251 (CB1 receptor antagonist), AM630 (CB2 receptor antagonist),capsazepine (TRPV1 antagonist), GW9662 (PPARγ antagonist), GW6471 (PPARα antagon-ist), and O-1918 (proposed endothelial cannabinoid receptor antagonist). Data were analysedby one-way ANOVA and Dunnett's post hoc test, and are reported as mean ± S.E.M. 50μM EDTA caused a drop of TEER of about 20% (indicating an increase in cell permeability).Application of phytocannabinoids caused a more rapid recovery of TEER than observed inthe control group in a concentration dependent manner (area under the curve (AUC); vehicle3728 +/- 234; 10 μM delta-9-tetrahydrocannabinol (THC) 2148 +/- 475, P<0.05; 10 μMcannabidiol (CBD) 1201 +/- 53, P<0.01). By contrast, application of endocannabinoidscaused a further and sustained drop in TEER in addition to the effects of EDTA (AUC; vehicle3553 +/- 211; 10 μM anandamide 4985 +/- 409, P <0.05; 10 μM 2-arachidonoylglycerol (2-AG) 5373 +/- 271, P <0.05). Only CB1 receptor antagonism inhibited the actions of cannabi-noids in Caco-2 cells (THC P <0.001; CBD P <0.001, anandamide P<0.001; 2-AG P <0.05).These findings suggest that endocannabinoids may play a role in modulating intestinalpermeability, and that phytocannabinoids may have therapeutic potential in the reversal ofthe abnormally permeable intestine in shock and sepsis.

989

Pre-Operative Nomogram to Predict Risk of Peri-Operative Mortality FollowingHepatic Resections for MalignancyMashaal Dhir, Smith Lynette, Fred Ullrich, Leiphrakpam Premila, Quan Ly, Aaron R.Sasson, Chandrakanth Are

INTRODUCTION: The majority of hepatic resections for malignancy are performed in olderpatients with major co-morbidties. There is currently no pre-operative, patient-specificmethod to determine the likely peri-operative mortality for each individual patient. The aimof this study was to develop a pre-operative nomogram based on the presence of co-morbdities to predict risk of peri-operative mortality following hepatic resection for malig-nancy. METHODS: The National Inpatient Sample database was queried to identify adultpatients that underwent hepatic resection for malignancy. The pre-operative co-morbidties,identified as predictors were used and a nomogram was created with multivariate regressionusing the Taylor expansion method in SAS Software, SURVEYLOGISTIC Procedure. SampleA (2000- 2004) was utilized to develop the model and Sample B (2005) was utilized tovalidate this model. RESULTS: A total of 4405 and 1072 patients were included in SamplesA and B. The overall actual observed peri-operative mortality rate for Samples A and B was4.2% and 3.5% respectively. The decile- based calibration plot for Sample A revealed excellentagreement between the observed probabilities and nomogram predicted probabilities. Sim-ilarly, the quartile-based calibration plot for Sample B revealed good agreement betweenthe observed and predicted probabilities. The accuracy of the nomogram was further rein-forced by a good concordance index of 0.80 with a 95% confidence interval of 0.73 to 0.86.CONCLUSION: This pre-operative nomogram has been shown to accurately predict therisk of peri-operative mortality following hepatic resection for malignancy.

S-865 SSAT Abstracts

990

Intestine-Specific Deletion of p38α MAPK Perturbs Enterocyte KineticsFollowing Massive Small Bowel ResectionDerek Wakeman, Jennifer A. Leinicke, Mark E. McMellen, Jun Guo, Christopher R.Erwin, Brad W. Warner

Introduction: Intestinal adaptation after massive small bowel resection (SBR) is characterizedby enhanced crypt cell proliferation and apoptosis. Both of these responses are modulatedby epidermal growth factor receptor (EGFR) signaling. We have demonstrated In Vitro thatp38α MAPK is required for apoptosis mediated by EGFR inhibition. The purpose of thisstudy was to determine the effect of In Vivo deletion of p38α expression on intestinaladaptation after SBR. Specifically, we tested the hypothesis that resection-induced enterocyteapoptosis would be altered by p38α MAPK deletion. Methods: Inducible, intestine-specificp38αMAPK-null mice (Ip38-null) were created by crossing mice with a tamoxifen-inducible-Cre-fusion protein under control of the villin promoter with mice in which the floxed p38αMAPK gene had been tagged for recombination. Tamoxifen was injected intraperitoneallyat a dose of 0.5mg/day for 2 days. Seven days after first injection Ip38-null mice underwenteither 50% SBR (n=10) or sham operation (n=9). Wild-type littermate controls (n=7) werealso injected with tamoxifen and subjected to 50% SBR. Mice were sacrificed after 3 daysand the remnant bowel was harvested for analysis. Crypt depth, villus length, crypt cellproliferative and apoptotic rates were measured. Results: Western blot analysis confirmedthat Ip38-null mice had a >90% reduction in p38α MAPK protein levels. After 50% SBRIp38-null mice exhibited significant adaptive increases in crypt depth, villus length, prolifera-tion, and apoptosis relative to shams. After SBR crypt and villus elongation in the Ip38-nullmice were comparable to the adaptive changes in the wild-type controls. However, afterSBR the Ip38-null mice had slightly increased rates of crypt proliferation relative to wild-type controls (17.3 +/- 0.8 vs. 15.2 +/- 0.8 BrdU labeled cells/crypt, p=.09). In addition,Ip38-null mice had lower apoptotic rates in the crypt compartment compared to wild-typecontrols after SBR (6.85 +/- 0.6 vs. 10.9 +/- 0.8 apoptotic bodies/100 crypts, p<.005).Conclusions: Deleting p38αMAPK in an inducible, intestine-specific manner does not impairintestinal adaptation in mice. However, loss of p38α MAPK does partially block the increasein crypt apoptosis normally seen after SBR. These findings support an important role forp38α MAPK in the kinetics of enterocyte turnover after small bowel resection.

991

Mucosa-Associated Microbiota and Risk of Chronic and/or Relapsing PouchitisAfter Restorative Proctocolectomy for Ulcerative ColitisMarco Scarpa, Alessia R. Grillo, Anna Pozza, Diego Faggian, Cesare Ruffolo, MelaniaScarpa, Renata D'Incà, Mario Plebani, Giacomo C. Sturniolo, Ignazio Castagliuolo, ImerioAngriman

Background Pouchitis can be classified as occasional or relapsing and about 5% of patientsdevelops a chronic inflammatory condition. Chronic pouchitis can lead to the pouch failure.Aims of our study were to identify possible relationship between chronic/relapsing pouchitisand mucosa-associated microbiota, systemic and local inflammation, local cytokines networkand toll like receptors expression. Patients andmethods In this prospective study 32 consecut-ive patients who underwent restorative proctocolectomy, coming for follow up endoscopyin our out-patient department were recruited in 2007. Clinical disease activity was classifiedusing Pouchitis Disease Activity Index. During pouch endoscopy biopsies from the ilealpouch were obtained to culture bacteria adherent to the mucosa and perform routinehistology. Systemic inflammatory status was graded according to blood cell count, ESR,CRP, and albuminemia. Local inflammatory status was assessed with faecal lactoferrin levelsanalyzed by quantitative ELISA. Serum and mucosal levels of IL-1β, IL-6 and TNF-α weremeasured with immunometric assays. The mucosal expression of toll-like receptors forbacterial lipopolysaccharide (TLR4) and peptidoglycan (TLR2) were measured by quantitativeReal Time RT-PCR. After a median follow up of 23 (IQR 20-24) months the patients werecontacted for a reassessment of current and past disease activity. Univariate logistic regression,non parametric statistics and survival analysis were performed. Results Clinical diagnosis ofpouchitis (PDAI>7) had been made in 10 patients in 2007. During the follow up 3 of thesepatients failed to recover and thus had chronic pouchitis while other 3 of them had relapsingepisodes of pouchitis. In patients with chronic/relapsing pouchitis mucosal TLR2 and TLR4were more expressed than in those with no or only one episode of pouchitis (p=0.036 andp=0.016, respectively). The number of CFU of mucosa-associated Clostridiaceae spp washigher in patients with chronic/relapsing pouchitis than in those with no or only one episodeof pouchitis (p=0.031). The presence of Clostridiaceae spp adherent to the ileal mucosa wasassociated to a significant risk of chronic/relapsing pouchitis [OR: 14 (95% CI 0.887-224.021), p=0.045]. Survival analysis showed that the presence of Clostridiaceae spp was

SS

AT

Ab

stra

cts