SS

AT

Ab

stra

cts

following laparotomy. Subgroups were vagotomized 3-4 days prior to experiments or receivedpharmacological inhibition of the acetylcholine α7 subunit with the inhibitor α-bungarotox-ine (1μg/kg i.p. 1h, 3h, 9h prior to ileus induction each n=6), while control animalswere sham operated and remained otherwise untreated. Three hours after small bowelmanipulation a 2 cm jejunal segment was harvested with the mesentery attached. Mesentericafferent nerve discharge was recorded In Vitro generating a multi-unit signal with subsequentcomputerized analysis including recording of intestinal motility. Afferent nerve discharge atbaseline, responses to chemical stimulation with bradykinin (0.5 μM), 5-HT (500 μM) andmechanical stimulation by continuous ramp distension to 60 mmHg were studied. Data aremean ± SEM. Results: Peak amplitudes of intestinal motor events were 0.70±0.02, 0.69±0.04,and 0.67±0.03 mmHg and afferent nerve discharge was 12±2, 11±1, 14±1 1 imp sec-1following vagotomy, α-bungarotoxine, or sham operation, respectively with no differencebetween groups (all n.s.). Increase of afferent discharge to 5-HT was 6 ± 1 imp sec-1 abovebaseline following α-bungarotoxine which was similar compared to 8 ± 1 imp sec-1 in shamcontrols, while the response was reduced to 0.7±1.6 imp sec-1 in chronically vagotomizedanimals (p<0.05). Bradykinin was followed by 20±2, 19±1 and 22±1 imp sec-1 after vago-tomy, α-bungarotoxine or sham operation respectively, while peak firing rate was 83±9,95±12 and 80±8 imp sec-1 during intraluminal ramp distension at 60 mmHg (all n.s.). Atluminal distension from 10 to 30mmHg, afferent discharge was lower in vagotomized animalscompared to sham controls (p<0.05), but unchanged after α-bungarotoxine. Conclusions:Sensitivity to 5-HT and low-threshold distension is mediated via vagal afferents duringpostoperative ileus, while sensitivity to high threshold distension and bradykinin is independ-ent of intact vagal afferents. Inhibition of intestinal motility early, i.e. 3 hours after inductionof postoperative ileus, does not appear to depend on vagal innervation.

970

Eating Behavior in Rats Subject to Vagotomy, Sleeve Gastrectomy andDuodenal SwitchYosuke Kodama, Chun-Mei Zhao, Baard Kulseng, Duan Chen

Background/aim: Obesity is a multifactorial disease and the treatments include dieting,exercises, drugs and various surgical procedures. Recently, we reported that gastric bypasssurgery caused body weight loss without reducing food intake and that high-fat diet-inducedobesity was associated with increased calories per meal but not per day in rats. In the presentstudy, we examined the food intake, eating behavior and metabolic parameters in ratsunderwent bilateral truncal vagotomy, sleeve gastrectomy and duodenal switch procedures.Methods: Body weight (BW) was recorded weekly throughout the period of the study. Thefood intake, eating behavior and metabolic parameters were measured pre- and 2 or 9 weekspost-operatively by a comprehensive laboratory animal monitoring system (the so-calledCLAMS). Adult rats were subjected to bilateral truncal vagotomy plus pyloroplasty to preventgastroparesis (VTPP), pyloroplasty alone (PP) or sham operation as controls. 10 weeks aftercompleting CLAMSmeasurements, sleeve gastrectomy (SG) or duodenal switch (DS)(withoutSG) was performed in sham- or PP-operated rats, respectively, and the same CLAMSmeasure-ments were repeated as before. Afterwards, the SG-rats were subjected to DS and the VTPP-rats were subjected to both SG and DS simultaneously. Results: Survival rates were 100%for sham- (7/7), PP- (7/7) and VTPP-operations (7/7), 86% for SG (6/7), 71-83% for DSalone (5/7) or DS with SG (5/6) that was performed early, and 14% for SG+DS (1/7) thatwere performed at the same time. VTPP reduced BW (10%) transiently (1 week post-operatively), while PP- or sham-operation was without the effect. SG caused a reduced BW(10%) for 6 weeks, while DS alone or SG followed by DS led to a continuous BW loss from15% at 1 week to 50% at 9 weeks postoperatively. Food intake was higher and satiety ratiowas lower during nighttime than daytime in all groups of surgeries. VTPP was without anymeasurable effects on food intake, eating behavior and metabolic parameters. SG increaseddrinking activity and energy expenditure but was without the effects on food intake andeating behavior. DS regardless of accompanying with SG reduced food intake by about 60%during nighttime but not daytime, and did not affect energy expenditure. Conclusions:Weight loss after VTPP, SG or DS differed in terms of degree, duration and underlyingmechanisms. DS without SG was most effective in long-term, at least partly due to thereduced food intake. In addition to food intake and eating behavior, possible involvementsof absorption, gut hormones and the brain-gut axis need to be further studied.

971

Perioperative Anti-Tumor Necrosis Factor Alpha Agnents Do Not Increase theRate of Early Postoperative Complications in Crohn's DiseaseBasil S. Nasir, Eric J. Dozois, Robert R. Cima, John H. Pemberton, Bruce G. Wolff,William J. Sandborn, Edward V. Loftus, David W. Larson

BACKGROUND: There have been numerous studies with conflicting results regarding theuse of anti-tumor necrosis factor (TNF) alpha and its relationship to postoperative outcome.The aim of our study was to examine the rate of postoperative morbidity in patients receivinganti TNF alpha in the perioperative period. METHODS: All patients undergoing surgery forCrohn's disease from 2005 till 2008 were abstracted from a prospective database. Patientswho underwent surgery which included a suture or staple line at risk for leaking wereselected for the study. A retrospective review of medical records was performed. The studygroup comprised patients treated with anti-TNF alpha within 8 weeks preoperatively or upto 30 days postoperatively. The remainder of the patients did not receive the drug in thattime period. Patient characteristics, disease severity, medication use, operative interventionand 30-day complication were compared between the two groups. RESULTS: Three hundredand seventy patients were selected for analysis in this study, of which 119 received the drugwithin the allotted time period and 251 did not. The groups were similar in baselinecharacteristics, perioperative risk factors and procedures. The study group had more severedisease overall as measured by the American College of Gastroenterology (ACG) categoriesof disease (50% severe fulminant disease in the study group versus 18% in the unexposedgroup, p < 0.001). There was no significant association of anti-TNF alpha therapy and anypostoperative complications (27.9% in study group versus 30.1% in unexposed group, p =0.63) nor intra-abdominal infective complications (5.0% in study group versus 7.2% inunexposed group, p = 0.44). Univariate analysis showed that the only factors associated

S-862SSAT Abstracts

with an increase in postoperative intra-abdominal infections were age and penetrating disease.CONCLUSIONS: The use of anti-TNF alpha in the perioperative period is safe and doesnot seem to be associated with an increase in overall or infective complications for patientswith Crohn's disease undergoing surgery.

972

The Relationship Between Pre-Operative Comorbidity, the SystemicInflammatory Response and Survival in Patients Undergoing Surgery forColorectal CancerCampbell S. Roxburgh, Jonathan J. Platt, Fiona Leitch, Paul G. Horgan, Donald C.McMillan

Introduction: Besides tumor characteristics, colorectal cancer progression and survival isalso determined by host factors, in particular a systemic inflammatory response (GlasgowPrognostic Score/GPS). The basis of this stage independent relationship with survival isunclear, however pre-op systemic inflammation may reflect comorbidity. Indeed, validatedscores such as elevated Charlson comorbidity index (CCI), National Institute on Aging/National Cancer Institute (NIA/NCI) Index and Adult Comorbidity Evaluation-27 (ACE-27)are related to poor colorectal cancer survival. This study examines the relationships betweenpre-op comorbidity, systemic inflammation (GPS) and survival in colorectal cancer. Methods:Patients having elective curative resection were studied (n=302,1997-2005). A pre-op C-reactive protein>10mg/L and albumin<35g/dl each score 1 and the GPS is constructed (GPS0, 1 or 2). Comorbidity data was abstracted from casenotes review. Patients were classifiedby 4 separate scores; the CCI, NIA/NCI index, ACE-27 and Lee Cardiac Risk Index (LCRI).Deprivation category, smoking status and body mass index (BMI) was collected. Results:Most were >65yrs (66%), Stage I/II disease (60%), a normal GPS (62%) and had a lowcomorbidty burden. Median follow-up was 74 months during which 135 died (85 fromcancer). On multivariate analysis for cancer survival, age (HR 1.30,P=0.057), TNM stage(HR 2.73,P<0.001), LCRI (HR 1.38,P=0.014) and GPS (HR 1.85,P<0.001) were independ-ently related. On multivariate analysis for overall survival, age (HR 1.50,P<0.001), TNMstage (HR 1.84,P<0.001), ACE-27 (HR 1.31,P=0.005) and the GPS (HR 1.61,P<0.001) wereindependently related. Results were similar in node negative disease (n=180). Old age wasrelated to increasing comorbidity (ACE-27, CCI, LCRI (all P<0.005)) and elevated GPS(P<0.005). High BMI was related to higher comorbidity assessed with CCI, ACE-27 andNCI/NIA scores (all P<0.01). Smoking history and deprivation were related to increasingburden of comorbidity (all P<0.05). GPS had a weak association with comorbidity burdenassessed with the ACE-27 (P=0.065), CCI (P=0.016), LCRI (P=0.095) and the NIA/NCIindex (P=0.084). Discussion: Pre-op comorbidity measures, particularly Lee cardiac riskindex are important indicators of cancer survival. Generalised comorbidity does not explainthe relationship between the systemic inflammatory response and survival. The tumor andits response to therapy form themainstay of current cancer reatment, however it is increasinglyapparent novel “host-related” targets may be important. These include attenuation of thehost inflammatory response and optimisation of host physiology.

973

Relationship of EMAST and Microsatellite Instability Among Patients WithRectal CancerBikash Devaraj, Sonia Ramamoorthy, Robert S. Sandler, Temitope O. Keku, Aaron Lee,Linda Luo, Kathie McGuire, Betty L. Cabrera, Katsumi Miyai, John M. Carethers

Background. Elevated microsatellite instability at selected tetranucleotide repeats (EMAST)is a genetic signature identified in two-thirds of sporadic colon cancers (and other cancers)and is associated with heterogeneous expression of the DNA mismatch repair (MMR) proteinhMSH3. Unlike microsatellite instability-high (MSI) in which hypermethlation of hMLH1occurs followed by multiple susceptible gene mutations, EMAST may be associated withinflammation and relaxation of MMR function with the subsequent biological consequencesnot known. We evaluated EMAST and MSI in a population-based cohort of rectal cancers,as this has not been previously determined. Methods. We analyzed 147 sporadic cases ofrectal cancer using 5 tetranucleotide microsatellite markers and NCI-recommended MSI(mono and dinucleotide) markers. EMAST and MSI determinations were made on analysisof DNA sequences of the PCR products, and determined positive if at least 2 loci were foundto have frameshifted repeats upon comparison between normal and cancer samples fromthe same patient. We correlated EMAST data with race, gender, and tumor stage, andexamined a subset of samples for lymphocyte infiltration. Results. Among this cohort ofpatients with rectal cancer (mean age 62.2 +/-10.3 years, 36% female, 24% African American),3/147 (2%) showed MSI (3 males, 2 African American) and 49/147 (33%) demonstratedEMAST (frequency of markers positive are in Table 1). Rectal tumors from African Americanswere more likely to show EMAST than Caucasians (17/36, 47% vs 30/105, 29%, P=0.04),and approached an association with advanced stage (18/29, 62% vs local disease 11/29,38%, P=0.06). There was no association between EMAST and gender (female 19/53, 36%vs male 28/88, 32% P>0.05). In the subset of tumors analyzed for lymphocytes histologically,EMAST was more prevalent in rectal tumors that showed peri- or intra-tumoral infiltrationcompared to those without. Conclusions. The frequency of EMAST in rectal cancers (~33%)is less than reported for colon cancer (~66%), and MSI is rare. EMAST is associated withAfrican American race and may be more commonly seen with metastatic disease. The etiologyand consequences of EMAST is under investigation, but its association with immune cellinfiltration suggests inflammation may play a role for its development.Frequency of EMAST Marker Mutations