fullness (p<0.01), epigastric pain (p<0.01) and burning (p<0.01). Improvements in PedsQLand SMW total scores remained when analysis was limited to normal or delayed gastricemptying (p<0.05, p<0.05). Only 13% (3/24) needed tube feeds and 13% (3/24) parenteralnutrition after GES. School absences decreased from 57% to 31% of school days. Overall,65% (13/20) reported their health was much improved after GES versus 15% (3/20) thesame or worse. The majority (15/20) were satisfied with GES. Three were not satisfied dueto lack of improvement, one developed back pain and another was later diagnosed with aneating disorder. Five reported complications. Four had discomfort or tenderness at theimplantation site and another had a dead battery. Conclusions: In the largest series to dateof pediatric patients who have undergone GES for GP and/or FD, we have found significantand sustained improvement not only in upper GI symptoms but also in quality of life andperception of global health. Patients were less dependent on tube feeding or parenteralnutrition and had fewer school absences. The majority is satisfied with the decision to placeGES. Future studies are needed to assess for possible placebo effect and to evaluate predictorsof outcome and long-term prognosis.

1086

Gastroparesis in the US - an Emerging EpidemicSravanthi Parasa, Arun Raghav Mahankali Sridhar, Kevin Olden

OBJECTIVES/HYPOTHESIS: The epidemiology of gastroparesis is largely unknown. Thepast decade has witnessed little progress in the management of gastroparesis. We sought tocharacterize the time trends in the proportion of patients hospitalized with gastroparesis.STUDY DESIGN: Retrospective cross-sectional study. METHODS: Using the NationwideInpatient Sample database, temporal trends in hospitalisations for gastroparesis was per-formed from years 1994 to 2009. All patients admitted to hospital between 1994 and2009 with a primary discharge diagnosis of Gastroparesis, identified by the InternationalClassification of Diseases, Ninth Revision procedure codes were included (536.3). Temporaltrends in the change in the hospital charges, length of stay, proportion of Emergencydepartment visits admitted as inpatients over the selected time period were evaluated.Statistical analysis was done using STATA 12.0 MP after applying appropriate weights forthe propensity score as previously determined by AHRQ recommendations. RESULTS:Between 1994 and 2009, the absolute number of Gastroparesis hospitalizations increasedfrom 918 to 16738, (p<0.01). There was a significant increase in charges from $14571 to$29481 in 2009 (p<0.05). Also of note there has been a linear increase in the proportionof patients admitted to inpatient service from the emergency department - 43% to 65%(p<0.05).( See figure) CONCLUSIONS: There was a concomitant significant increase ofhospital cost for each gastroparesis hospitalization along with increase in the number ofpatients being admitted to hospital. This represents a significantly increase in the costsincurred by this disorder in the past 15 years and needs to be further studied.

1087

Gastric Interstitial Cells of Cajal Quantification on Full Thickness GastricBiopsy Provides Prognostic Information for Treatment Responses to GastricElectric StimulationAbhinav Sankineni, Sean Harbison, Rebecca Thomas, Henry P. Parkman

Patients with gastroparesismay have dysfunction or paucity of enteric nerves and/or interstitialcells of Cajal (ICC) that may contribute to pathophysiology, symptoms, and treatmentoutcomes. AIM: To evaluate ICC and innervation of the enteric nervous system in patientswith gastroparesis and to determine if these influence treatment outcomes with gastric electricstimulation (GES). METHODS: Full thickness gastric biopsies were obtained from patientswith gastroparesis undergoing surgical placement of a gastric electric stimulator. Biopsieswere obtained from the anterior gastric body using 35 mm gastrointestinal stapler (EthiconTA35). Biopsies were placed in formalin, followed by paraffin embedding, sectioning, andimmunohistochemical staining for the presence of nerves with neuron specific enolase (NSEMonoclonal Antibody; DAKO, Carpinteria, CA) and for ICC (c-Kit Antibody; Ventana,Tucson, AZ). The number of ganglia, ganglion cells, and ICC were counted on 5 consecutivehigh power fields (400X). Patients were followed over time and classified as responders ornon-responders to GES using the modified Clinical Patient Grading Assessment Scale(CPGAS): 0=no change, +7=completely better. RESULTS: 74 patients (average age 38.1±11.7years; 61 females) with refractory gastroparesis (35 diabetic, 35 idiopathic, 4 post Nissen)underwent implantation of a gastric electric stimulator from July 2010 to November 2011.Of the 65 patients who had full thickness biopsy specimens, 45 improved with GES whereas19 patients stayed the same or worsened. The number of ICC/hpf were decreased in patientswho did not improve with GES compared to those who improved: ICC in outer longitudinalmuscle layer [0.82±0.11 vs 1.42±0.15; p=0.002]; inner circular muscle layer [2.20±0.19 vs2.86±1.49; p=0.028]; and myenteric plexus [0.91±0.16 vs 1.02±0.09; p=0.54]. The CPGASimprovement score at follow-up (average 6.5 months) correlated with the number of ICCin the outer longitudinal muscle layer (r=0.272; p=0.003), the inner circular muscle layer(r=0.210; p=0.021), and myenteric plexus (r=0.181; p=0.057). The number of ganglioncells (4.20±0.69 vs 3.95±0.34) and number of ganglia (1.07±0.14 vs 1.06±0.06) were similar

S-195 AGA Abstracts

in patients not improving and those that did improve. Although patients that improvedwere older (39.7±11.3 vs 33.6±11.3 yrs; p=0.05) and tended to have diabetes (78.8 vs62.5%; p=0.149), age and diabetes were not associated with changes in ICC, ganglia, organglion cells. CONCLUSIONS: This histologic study of gastroparesis patients with refract-ory symptoms shows that nonresponders to GES had lower number of ICC than patientsthat improved. In contrast, the ganglia and ganglion cells quantitatively appeared similar.Thus, information from gastric full thickness specimens may provide prognostic informationon the outcome of GES treatment for gastroparesis.

1088

Outcomes and Predictors of Improvement in Patients With GastroparesisFollowed Prospectively for 48 WeeksPankaj J. Pasricha, Katherine P. Yates, James Tonascia, Linda Anh B. Nguyen, Henry P.Parkman, Gianrico Farrugia, Kenneth L. Koch, Aynur Unalp, William J. Snape, JorgeCalles, William L. Hasler, Thomas L. Abell, Richard McCallum, Irene Sarosiek, Linda A.Lee, Frank A. Hamilton

Background. The management of patients with gastroparesis remains a therapeutic challengeowing to the heterogeneity of the patient population and multiple mechanisms that contributeto symptom generation. Little is known about the clinical or pathophysiological character-istics, if any, which may be helpful in predicting response over time. In this study, our aimwas to determine if demographic characteristics, pattern of gastric emptying, etiology ofgastroparesis, symptom nature and severity are predictive of improvement. Methods . Westudied 357 patients with gastroparesis (including idiopathic as well as that associated withboth type 1 and type 2 diabetes mellitus) enrolled in the NIDDK Gastroparesis ClinicalResearch Consortium (GpCRC) registry, and who had completed 48 weeks of follow-up.These patients were seen every 16 weeks and prescribed medications or other therapies asper the judgment of the treating physician. Improvement was defined as a decline of at least1 point in the GCSI scores at 48 weeks compared to the baseline score. P values weredetermined from a multiple logistic regression analyses of improvement using multiplebaseline parameters as putative predictors of improvement. Results. Of 357 patients, 99(28%) reported GCSI scores consistent with improvement as defined above. The followingbaseline attributes correlated significantly with GCSI improvement at 48 weeks: Age > 50years (Odds Ratio, OR=1.65 versus younger patients; P=0.04); smoking history (ever versusnever; OR=0.49; P=0.01); weight (BMI>25 versus <25; OR= 0.54; P=0.01); initial infectiousprodrome (OR=1.88; P=0.03); depression (Beck Depression Inventory>20 versus <20; OR=0.57; P=0.02). On the other hand, gender, race (white versus other), etiology (idiopathic,type 1 or type 2 DM), presence or absence of severe abdominal pain (PAGI-SYM symptomscore >3), presence or absence of inflammation (CRP>1 or ESR>20), degree of gastricretention or the PAGI quality of life score at baseline did not correlate with improvement.Conclusions: After nearly a year of structured follow-up and treatment at centers of expertise,only a minority of patients report symptomatic improvement. Older patients and those withan infectious prodrome have a greater probability of improving while patients who areoverweight, depressed or with a smoking history have a lower probability. These resultsemphasize the chronic nature of gastroparesis and lead to interesting hypotheses aboutinterventions directed against risk factors for non-improvement.

1089

Dietary Oligofructose Prevents Development of Endotoxemia-Induced LeptinResistance and Hyperphagia Induced by High Fat Feeding in Rats: Mediationby a GLP-2 Dependent MechanismPornchai Leelasinjaroen, Shi-Yi Zhou, Yuanxu Lu, Xiaoyin Wu, Phattrawan Pisuchpen,Yundong Sun, John Y. Kao, Chung Owyang

High-fat feeding leads to leptin resistance, hyperphagia and development of metabolicsyndrome and obesity. This appears to be mediated by changes in gut microbiota causingan increase in intestinal permeability and metabolic endotoxemia. Previous studies reportedthat prebiotic carbohydrates stimulate GLP-2 production in the proximal colon and improvegut barrier functions. We hypothesize dietary oligofructose should prevent the developmentof endotoxemia-induced leptin resistance and hyperphagia induced by high fat feeding. Totest this, we performed studies on rat fed with a high fat diet (HFD, 53% kcal/g fat) withor without supplement with the fermentable non-digestible oligofructose (10%). Rats fedwith 8 wks HFD gained 12%more weight and a 78% increase in visceral adipocyte comparedto controls given a regular diet (11.3% kcal/g fat) (439+5 g vs 392+3 g, n=4, P<0.05). HFDcaused intestinal inflammation characterized by a 2.3, 4.2 and 8.3 fold increase in INFγ,IL-10 and IL-1β in the intestinal mucosa. This was accompanied by an 8 fold increase ingut permeability In Vivo (0.04 vs 0.32 μg/ml, 4000 Da fluorescent dextran-FITC) and a 19%decrease in transepithelial electrical resistance in the proximal colon mucosa. HFD increasedplasma LPS levels from 0.35 to 1.61 EU/ml (P<0.05). Dietary supplement with oligofructoseprevented the development of endotoxemia and adipocyte proliferation. These were accom-panied by increased synthesis of tight junction protein ZO1, normalization of gut permeabilityand marked reduction of intestinal inflammation. Oligofructose also enhanced GLP2 expres-sion in the proximal colon by 21% compared to controls. In Vitro studies showed GLP2 (1μM) caused a 2 fold increase in occludin expression in colonic explants cultured for 24 hr.To provide direct evidence that endotoxemia alters vagal sensory inputs, neuronal dischargesof single nodose ganglia innervating the proximal gut were recorded. Administration ofleptin (225 pmol) caused an increase in neural discharge (16.4+4 impulse/20 sec) in ratsfed a regular diet. Neuronal responses to leptin were reduced to 5.4+1.2 impulses/20 secin rats fed with HFD (P<0.05). In contrast, HFD rats treated with oligofructose showed anormal response (15.8+3.6 impulses/20 sec). The normalization of vagal responsiveness toleptin was accompanied by a reduction in feeding. Oligofructose reduced energy intake inrats fed with HFD (191 kcal/kg/day from 210 kcal/kg/d; P<0.05) to values similar to thosetaking a regular diet (185 kcal/kg/d). In conclusion, we showed that dietary oligofructoseimproves gut barrier function and prevents the development of endotoxemia-induced leptinresistance and hyperphagia induced by high fat feeding. These findings may have therapeuticsignificance in the prevention and/or treatment of metabolic disorders such as obesity andmetabolic syndrome.

AG

AA

bst

ract

s