‘trait’ psychological factors, somatic symptom reporting and quality of life in FD. Abuseexerts its effect directly on somatic symptom reporting, rather than indirectly throughpsychosocial factors. A reciprocal relationship between somatic symptom reporting and ‘statenegative affect’ was found. Gastric sensitivity influences somatic symptom reporting, butnot vice versa. Somatic symptom reporting and personality are the main determinants ofphysical and mental QoL, respectively.

Figure 1

495

Evaluation of Histologic Criteria for the Diagnosis of Gastroesophageal RefluxDiseaseMichael Vieth, Roberto Fiocca, Börje Wernersson, Lis Ohlsson, Magnus Ruth

Background. Standardized criteria for the histologic assessment of gastroesophageal refluxdisease (GERD) were recently developed and evaluated by an international working group,and showed promising levels of intra- and inter-assessor agreement1. To gain clinical accept-ance these criteria require appropriate validation2.Aim. To evaluate new criteria for histologicmarkers of GERD in a well characterized primary care population. Methods. The studyincluded primary care patients from the Diamond study3 (ClinicalTrials.gov: NCT00291746)who had not previously taken proton pump inhibitors and who had upper gastrointestinalsymptoms ≥ 2 times a week for ≥ 4 weeks, and at least mild symptoms on ≥ 3 days inthe week before biopsy. GERD was diagnosed as any of the following: Los Angeles gradereflux esophagitis; distal esophageal pH < 4 for > 5.5% of 24 hours; ≥ 95% symptomassociation probability. Biopsies were collected at 0.5 cm and 2.0 cm above the Z-line anddata were analyzed in a single-blind manner by one pathologist (MV). Variables assessedincluded total epithelial thickness (TET), papillary length (PL), basal cell thickness (all inμm), presence of dilated intercellular spaces (severity score: 0-2)1, number of inflammatorycells (in one high powered field), presence of necrosis, and presence of active or healederosions. Logistic regression analysis of the histologic variables was performed for GERDversus non-GERD. Cut-off values for sensitivity, specificity, and predictive power (positiveand negative) were calculated, and the receiver operating characteristic (ROC) curve wasanalyzed. Results. The analysis included 258 subjects: 138 (53%) GERD and 120 (47%)non-GERD. TET and PL (μm) at 0.5 cm and 2.0 cm were both significant predictors ofGERD (p < 0.005), as was the presence of eosinophils at 2.0 cm (p < 0.05). The optimalefficiency based on the ROC analysis was observed for TET at 0.5 cm, with sensitivity andspecificity values of 77% and 52%, and positive and negative predictive values of 65% and66%, based on a cut-off of 390 μm. TET and PL were significant (p < 0.05) predictors ofboth reflux esophagitis and of pathologic esophageal pH. TET and PL were also significantpredictors (p < 0.05) of non-erosive reflux disease versus non-GERD. Conclusions. Potentialnew criteria for the histologic evaluation of GERD showed good diagnostic properties whenassessed in a well-characterized primary care population with upper gastrointestinal symp-toms. References 1.Fiocca, R, et al. Hum Path 2010;41(2):223-31 2.Dent, J. Clin Gastroenteroland Hepatol 2007;5:4-16 3.Dent, J, et al. Gut 2010;59:714-21.

496

Common Cavity is a Marker of Proximal and Mixed Gastroesophageal Refluxand is More Frequently Associated With Reflux Episodes in Non-ErosiveReflux Disease Patients Than in Healthy VolunteersMentore Ribolsi, Paola Balestrieri, Maria Chiara Addarii, Sara Emerenziani, Michele Cicala

Background: Transient LES relaxations (TLESRs) are the major mechanism for the occurrenceof gastroesophageal reflux. The mechanisms underlying symptoms in non-erosive refluxdisease (NERD) patients are still unclear. Growing evidence shows that proximal extent andliquid-gas (mixed) composition of reflux episodes are relevant for symptom generationprobably through the involvement of both chemo- and mechano-receptors. The manometricphenomenon of common cavity (CC) is considered a marker of gastroesophageal reflux,although its specificity and relationship with the reflux pattern remain to be fully investigatedin GERD patients. High resolution manometry (HRM) combined with multichannel imped-ance (MI) monitoring allows accurate and simultaneous analysis of the reflux episodes andesophageal motility. Aim/Methods: To evaluate the relationship between reflux pattern andesophageal pressures during TLESRs, 12 NERD patients, following a 3-week pharmacologicalwashout, and 10 healthy volunteers (HV), sex and BMI matched, without evidence of hiatalhernia, underwent combined HRM-MI, before and 60 min., after a solid-liquid meal (1500Kcal) in a sitting position. A catheter with 36 solid state pressure sensors, located at 1 cmintervals, and with 9 impedance measuring segments placed between 5 and 21 cm abovethe LES, was used (MMS SOLAR HRIM, Enschede, The Netherlands). Reflux events weredefined and classified as mixed (liquid-gas) and proximal according to standardized criteria.A common cavity (CC) phenomenon was defined as a sudden rise in esophageal bodypressure to that of gastric pressure, associated with a reflux episode, maintained for ≥0.5

S-95 AGA Abstracts

sec. in ≥10 pressure tracings. TLESRs were defined according to standard criteria andagreement between 2 investigators was required. Results: 1 HV and 2 patients did notcomplete the test and were not analyzed. A total of 54 TLESRs occurred in the 9 HV (mean± SD, 6 ± 1.9). During TLESRs, in HV, 39 reflux episodes occurred (4.3 ± 1.4), of these23 (59%) were mixed and 21 (53%) proximal; 9 CCs were detected (23% of reflux episodes),7/9 associated with mixed and 8/9 with proximal reflux episodes. A total of 68 TLESRsoccurred in the 10 NERD patients (6.8 ± 4.9). During TLESRs, 62 reflux episodes (6.2 ±4.7) occurred, of these, 24 (38%) were mixed and 32 (52%) proximal; 24 CCs were detected(38% of reflux episodes), 18 associated with mixed (75%) and 20 (83%) with proximal refluxepisodes. Conclusions: During TLESRs, reflux episodes were more frequently associated withCC in patients than in HV. In both groups, CCs were strongly associated with liquid-gasreflux episodes and almost all of which with proximal reflux. The higher volume of mixedand proximal refluxes might be responsible for the esophageal distention during CC andmight be involved in symptom generation in NERD.

497

Assessment of Proton Pump Inhibitor Utilization After Negative MultichannelIntraluminal Impedance pH Testing for Refractory GERDJami A. Rothe, Angela Fought, Gabriela I. Escobar, Daniel Luger, Chang Lu, Peter J.Kahrilas, John E. Pandolfino

BACKGROUND/AIMS: Ambulatory reflux testing is used to evaluate GERD symptoms notresponding to PPI therapy. Negative endoscopy and reflux testing suggests an alternativediagnosis and patients may not require PPI therapy. The primary aim of this study was toassess the prevalence of PPI utilization in patients with negative multichannel intraluminalimpedance-pH (MII-pH) testing and determine predictors of continued PPI use. METHODS:Patients who had undergone MII-pH testing from 2006-2010 at a tertiary referral centerwere included in the study if they met three inclusion criteria: total acid exposure time<5%, total number of reflux events <73 and negative symptom index (SI). A retrospectivechart review was performed, eligible patients were contacted, and a cross-sectional telephonesurvey including questions about PPI utilization, health behaviors, demographics, symptomseverity (GERDQ) and quality of life (EQ-5D questionnaire) was administered. Physiologicvariables from MII-pH (total reflux episodes, acidic, weakly acidic, non-acidic events, percenttime pH<4, SI and SAP) were evaluated and compared. Statistical analysis was performedusing SAS v9.2. RESULTS: 87 patients met inclusion criteria and 60(69%) patients completedthe telephone survey. 33 had MII-pH testing/on PPI therapy and 27 had MII-pH/off PPItherapy. Of the 33 patients with MII-pH/on PPI, 27(81%) continued PPI therapy afternegative testing; only 26% of these patients had a secondary indication for PPI use (Barrett's4, PUD 3). Of the 27 patients studied with MII-pH/off PPI, 12(45%) continued PPI therapydespite a negative test and 17% had a secondary indication for PPI therapy. Over 80% ofpatients had a high disease impact score as measured by the GERDQ but only 32% ofpatients had high overall symptom scores (GERD Q≥8). The majority (>75%) of currentPPI users reported problems with anxiety (p=0.05) and pain (p=0.05) compared to non PPIusers as measured by the EQ-5D. Only a small minority of patients had used alternativetherapies (i.e. cognitive behavioral therapy, hypnotherapy). There were no significant differ-ences between PPI users and non-users when stratified by gender, age, BMI, race, maritalstatus, income, or education, physiologic variables, or PPI therapy at time of testing. CON-CLUSIONS: The majority of patients with a negative MII-pH study continued taking a PPIdespite not having a clear indication. Patients who continue PPI therapy are more likely tohave symptoms of anxiety or pain. The low overall GERQ score indicates a low likelihoodof GERD, but the high impact score across all patient groups reveals significant diseaseburden in patients with a negative MII-pH result. These results highlight the need forimproved strategies to identify alternative diagnoses and treatment strategies (i.e. othermedications, behavioral therapy) in PPI non-responders.

498

Squamo-Columnar Junction Locator Probe: From Bench to In-Vivo StudyYeong Yeh Lee, James Whiting, Elaine V. Robertson, John P. Seenan, Mohammad H.Derakhshan, Angela A. Wirz, Patricia Connolly, Kenneth E. McColl

Introduction The gastro-oesophageal junction is very mobile and constantly moving alongits caudal-craniad axis with breathing, swallowing and transient lower oesophageal sphincterrelaxation (TLESR). The only method currently available for studying its location is fluoro-scopic screening and this is limited by radiation exposure. We have developed a methodallowing continuous real-time monitoring of the squamo-columnar (SC) junction withoutradiation exposure. The method involves endoscopically clipping a small magnet (2mmdiameter and 1mm thick) to the SC junction and monitoring its position relative to theprobe comprising 26 Hall Effect sensors mounted at 5mm spacing on a circuit boardcontained within a silastic tube (diameter 3.2mm). The aims of the study were to assess theaccuracy of this technique on the bench and subsequently to validate the new techniqueagainst fluoroscopy. Methods For bench, the accuracy was determined by comparing therecorded position of the magnet along the length of the probe with its actual position alongthe probe. This was assessed with the magnet in various rotations and orientations bothanterior and posterior to the probe and various distances away from the probe. For In-Vivostudy, in eight healthy subjects, the magnet was attached and locator probe inserted. Duringsimultaneous fluoroscopy, subjects were asked to perform normal breathing, deep breathing,water swallows and finally advancement and retraction of locator probe over 12cm segment.The fluoroscopy recorded images at a rate of 5 frames per seconds and the locator at 8 Hz.The position recorded by fluoroscopy and locator at each second interval were comparedas well as amplitude of each complete manoeuvre. Results On bench, the accuracy of thesystem was related to the distance between magnet and probe, orientation of the field ofthe magnet relative to the probe and whether the magnet was positioned anterior, posterioror lateral to the probe. Over 95% of all studied orientations, an accuracy of less than ±10mmwas achieved up to a distance of 15mm between the magnet and the probe. In In-Vivostudy, the correlation co-efficient for all 224 position readings was 0.96 (95% confidenceinterval (CI) 0.89 - 0.96) and adjusted residual squared (R2) of 0.91. The amplitude forthe different manoeuvres was similar by the two techniques (see table). Conclusion On

AG

AA

bst

ract

s