AG

AA

bst

ract

sof hospital stay without recurrence of pain. A normal diet with a 30% lipidic content canbe safely used during the first days of refeeding.Comparison early vs. delayed oral refeeding

833

A Comparitive Study of Primary Conservative Treatment Versus Surgery forInfected Pancreatic Necrosis Over Ten Years (1997-2006)Manik Sharma, Pramod K. Garg, Debabrata Banerjee, Kaushal Madan, Peush Sahni,Girish K. Pande

BACKGROUND: Infected Pancreatic Necrosis (IPN) carries a high mortality. The standardtreatment in patients with IPN is surgical necrosectomy. Case series have shown successfuloutcome with conservative therapy. OBJECTIVE: To compare primary conservative treatmentwith primary surgical treatment for IPN. METHODS: All consecutive patients with acutepancreatitis (AP) were studied; those with IPN formed the study group. The diagnosis ofIPN was made if the pancreatic aspirate was culture positive or there was air in the pancreaticbed. Patients with IPN were divided into 2 groups based on two time periods i.e. 1997-2002 (group 1) and 2003-2006 (group 2). From January 1997 to December 2002, thetreatment protocol for IPN was surgical necrosectomy, drainage and lavage. Medical treatmentwas employed if surgery couldn't be done due to unacceptable surgical risk or patient'srefusal. From January 2003 to December 2006, the management for IPN was changed tomore conservative approach i.e. all patients were treated primarily with medical treatmentincluding antibiotics and if required percutaneous drainage. Surgery was carried out onlyif there was no improvement or deterioration as decided a priori.Mortality was the primaryoutcome measure. RESULTS: Of the 683 patients with AP, there were 312 patients ingroup 1 and 371 in group 2; necrotizing pancreatitis was present in 125 and 143 patientsrespectively. Of them, 30 in group 1 and 50 patients in group 2 had confirmed infectednecrosis. The treatment modality and outcomes of the patientswith IPN are given in theTable. Of the total 80 patients with IPN over 10 years, those treated medically had asignificantly better survival than those treated surgically (76.9% vs. 46.4%; p=0.005). CON-CLUSION: Adoption of primary conservative treatment resulted in a lower mortality, althoughstatistically insignificant, than primary surgical treatment in patients with IPN.Over the 10year period patients treated medically had a better outcome.Management of Infected Pancreatic Necrosis

*p=ns

834

A Comparison of Direct Endoscopic Necrosectomy and Usual TransmuralEndoscopic Techniques for the Treatment of Walled Off Pancreatic NecrosisTimothy B. Gardner, Prabhleen Chahal, Santhi Swaroop Vege, Bret T. Petersen,Christopher J. Gostout, Mark D. Topazian, Naoki Takahashi, Todd H. Baron

AIMS: Endoscopic transmural drainage of pancreatic necrosis has evolved from stent place-ment and irrigation to direct endoscopic necrosectomy. To evaluate the possible benefit ofdirect endoscopic necrosectomy, we compared this technique to standard endoscopic drain-age at our center. METHODS: All patients referred to Mayo Medical Center since April1998 for endoscopic drainage of walled off pancreatic necrosis (WOPN, formerly organizedpancreatic necrosis) were retrospectively identified. All patients had documented acutepancreatitis, pancreas-protocol CT scans showing WOPN, and all endoscopic proceduresperformed at our institution. Each patient underwent standard endoscopic cavity drainagebut were stratified into the direct endoscopic necrosectomy group if during any of theirendoscopic procedures, the necrotic cavity was directly entered with an endoscope anddebridement was performed - all others were in the standard drainage group. Success wasdefined as resolution of the necrotic cavity without surgical or percutaneous intervention.Two-tailed student's t-tests and Fisher's exact test were used to assess for differences betweengroups. The study was approved by the Mayo IRB. RESULTS: 45 patients were identifiedwho met study criteria - 25 underwent direct endoscopic necrosectomy and 20 underwentstandard endoscopic drainage. There were no differences in baseline patient or cavity charac-teristics. Successful cavity drainage was accomplished in 88% of those who underwentendoscopic necrosectomy vs. 45% for those receiving standard drainage (p<0.01) withoutan increase in the total number of procedures or in the usage frequency of an endoscopicallyplaced pancreatic irrigation tube. The maximum size of tract dilatation was larger in thedirect endoscopic necrosectomy group (16.8 mm vs. 14.2 mm, p=0.01). More patients inthe standard treatment group required surgical (30% vs. 4.0%, p<0.4) and percutaneousinterventions (20% vs. 0%, p<0.4). Time to resolution calculated from the time of initialdrainage to resolution by CT was similar between groups, although the standard treatment

T : 11501$$CH204-02-08 16:47:05 Page 120Layout: 11501B : e

A-120AGA Abstracts

group experienced a higher rate of recurrent collections within 6 months (40% vs 8%,p<0.03). Complications were limited to mild peri-procedural bleeding with equivalent ratesbetween groups. CONCLUSIONS: Direct endoscopic necrosectomy achieves higher rates ofsuccessful cavity resolution without a concomitant increase in the number of endoscopicprocedures, complication rate or time to resolution compared with standard endoscopicdrainage for the treatment of walled off pancreatic necrosis.

835

Assessment of Functional Brain Activation in Expectation of Visceral Pain in aRat Step-Down Passive Avoidance ModelZhuo Wang, Sylvie Bradesi, Jonathan R. Charles, Raina Pang, Jean Michel I. Maarek,Kevin Lee, Wendy J. Winchester, Emeran A. Mayer, Daniel P. Holschneider

BACKGROUND: Noxious visceral stimuli activate brain regions in the homeostatic afferentprocessing network as well as corticolimbic pontine modulatory networks. Aberrant modula-tion of visceral pain processing (“central pain amplification”) has been hypothesized as aneural correlate of visceral hypersensitivity, a cardinal symptom of functional gastrointestinaldisorders. An effective way to dissociate activation of the modulatory mechanisms fromafferent processing is to investigate brain responses in expectation of visceral pain. AIM: Todevelop an animal model to assess functional brain activation in expectation of visceral pain.METHOD: Male Wistar rats were trained with the step-down passive avoidance paradigmfor 18 trials/day for 2 days. All animals (n = 6 each group) were inserted with a colorectalballoon and placed on an elevated platform in an open arena. Step-down latencies weremeasured. Upon stepping down, experimental animals received a 60-mmHg, 20-s colorectaldistension (CRD) delivered through a barostat, whereas the balloon in controls remainedun-inflated. On Day 3, passive avoidance behavior was retrieved in 2 trials in the absenceof colorectal balloon. [14C]-iodoantipyrine was infused (i.v.) 45 s after the onset of the secondtrial, followed by immediate euthanasia. Brains were serially sectioned and autoradiographicimages were digitized and used to reconstruct 3-dimensional brains. Regional cerebral bloodflow related tissue radioactivity (rCBF-TR) was analyzed with statistical parametric mapping(SPM). RESULTS: Experimental rats learned to avoid CRD by remaining on the platform asevidenced by increased step-down latencies. The learned avoidance behavior was retrievedin the absence of balloon (35 ± 7 s vs. 5 ± 2 s in controls, Trial 1 on Day 3, mean ± SEM,P = 0.006). SPM analysis showed increases in rCBF bilaterally in the ventral aspect ofcingulate, anterior insular and orbital cortices, as well as the cerebellum, pons, and theposterior and ventral posterior thalamic nuclei in experimental rats. Decreases in blood flowwere found bilaterally in the piriform and visual cortices, subiculum, medial thalamus, andmedial and lateral hypothalamus. CONCLUSION: Rats were conditioned to noxious CRDin the step-down passive avoidance paradigm. In general agreement with human brainimaging findings, brain regions implicated in visceral pain processing (insula, thalamus) andmodulation (orbitofrontal cortex) were activated in expectation of CRD in the conditionedrats. This model may be useful in the evaluation of drugs aimed at central pain amplification.Supported by a grant from GalaxoSmithKline.

836

The Effects of Acute Esophageal Acid Exposure On the Expression Profile ofNeurotransmitter Receptor Genes in the Rat Cingulate CortexBanani Banerjee, Ivan M. Lang, Bidyut K. Medda, Robert Birkett, Reza Shaker

Exposure of the esophagus to HCl causes sensitization of neural activity identified by fMRIin the cingulate cortex in humans, but the mechanism of this central sensitization is unknown.Various neurotransmitters and their receptors have been reported to mediate excitatory,inhibitory or modulatory effects in pre or post synaptic neurons of the brain, but the roleof any of them in central sensitization due to esophageal acidification is unknown. AIMS:To determine whether and which neurotransmitter gene products change as a result of acuteacid exposure of the esophagus. METHODS: Six rats were anesthetized (Nembutol, 30mg/kg IP) and their mid esophagus perfused with either 0.1N HCl (N=3) or 0.1M PBS (N=3)for 20 min at 0.1ml/min. Four hours later the rats were anesthetized again (same dose),and the brain removed as quickly as possible and placed in a metal slicing block on a coldplate. The brain was then sliced sagitally at 2mm from midline. The cingulate cortex obtainedfrom the first slice on either side of midline was transected antero-posteriorly, frozen usingliquid nitrogen, and stored at -80oC. At a later date, the mRNA expression of five differentexcitatory neurotransmitter receptors: N-Methyl-D-Aspartate (NMDA) subtypes NR1, NR1splice variant NR1C2' and NR2B, serotonin 5HT2A receptor and Transient Receptor PotentialVanilloid 1 (TRPV1); and two inhibitory neurotransmitter receptors: GABA-A (γ-Aminobutyric Acid) and GABA-B were quantified from the anterior (ACC) and posterior (PCC)cingulate cortex samples of the HCl- and PBS-treated rats. Data were expressed as relativemRNA expression = 2-(Ct target-Ct β-actin). RESULTS: Among the genes evaluated, the relativeNR1C2' expression in ACC from HCl-treated rats was 20 fold higher compared to PBS-treated control rats (p< 0.001). Similarly, acute esophageal HCl exposure resulted in a 9fold increase in relative NR2B gene expression in the ACC compared to PBS-treated controlrats (p<0.002). Acute esophageal HCl exposure failed to cause any significant differencesin the expression pattern of examined neurotransmitter genes in the PCC. CONCLUSIONS:The upregulation of NMDA receptor subunits in the ACC suggests that acid exposure ofthe esophagus may induce plasticity in NMDA neurons of the ACC which may contribute tocentral sensitization and development of symptoms induced by gastroesophageal acid reflux.