21

BARRETT’S DYSPLASIA AND EARLY CANCER IN WOMEN —RESPONSE TO ENDOSCOPIC ABLATION WITHPHOTODYNAMIC THERAPYLois L. Hemminger, A.R.N.P., Herbert C. Wolfsen, M.D.*. Mayo Clinic,Jacksonville, FL.

Purpose: Barrett’s high-grade dysplasia (BE�HGD) and early adenocar-cinoma (ACA) are generally described as disease of Caucasian men.Relatively little information is available regarding this disease amongwomen and their response to endoscopic ablative therapy.Methods: Since 1997, 70 patients have undergone photodynamic therapy(PDT) for BE�HGD or ACA. Prior to treatment, all patients were evalu-ated with computed tomography, endosonography and surgery consulta-tion. All patients either declined, or were considered unfit for, surgery. PDTused 2mg/kg porfimer sodium IV with 630nm red laser light doses of150–250 J/cm fiber length. No balloon fiber-centering device was used andpatients received only a single course of PDT. Residual glandular mucosaremaining after PDT was destroyed using argon beam coagulation. Patientrecords were reviewed and treatment outcomes compared among male andfemale patients.Results: Women comprised 24% of patients in this series (17/70) including14 BE�HGD patients and 3 ACA patients. Two-sided Wilcoxon rank sumtesting found no statistically significant difference between these womenand their male counterparts with respect to age (median 69 years), Barrett’ssegment length (median 5 cm), effectiveness of complete ablation afterPDT only (59%), occurrence of stricture after PDT requiring dilation(20%), recurrence of BE�HGD or ACA after compete ablation (0%) andfollow up time (median 23 months).Conclusions: While BE�HGD and ACA may be found most frequently inmen, these are diseases that also affect significant numbers of women.Barrett’s screening and surveillance studies should emphasize the enroll-ment of adequate numbers of women. This study indicates that women andmen with BE�HGD and ACA respond equally well to endoscopic ablationtherapy.

22

PHOTODYNAMIC THERAPY AND ENDOSCOPIC MUCOSALRESECTION FOR BARRETT’S DYSPLASIA AND EARLYCANCERHerbert C. Wolfsen, M.D.*, Lois L. Hemminger, A.R.N.P.,Massimo Raimondo, M.D., Timothy A. Woodward, M.D. Mayo Clinic,Jacksonville, FL.

Purpose: Endoscopic mucosal resection (EMR) and endoscopic ablationwith porfimer sodium photodynamic therapy (PDT) has recently beencombined to improve the accuracy of histologic staging and remove su-perficial carcinomas. We report our experience using these techniques inpatients with Barrett’s high-grade dysplasia (BE�HGD).Methods: All BE�HGD or early cancer patients referred for endoscopicablation were evaluated with computed tomography and endosonography(EUS). All patients either declined, or were considered unfit for, surgery.EUS detected nodular or irregular folds within the segment of glandularmucosa in 3 BE�HGD patients. EMR using the ‘band and snare’ techniquewas performed followed 6 weeks later by endoscopic ablation with PDTwith 2mg/kg IV porfimer sodium and light doses of 175–250 J/cm fiberlength (Mayo Clin Proc, Nov 2002).Results: The only complications of EMR and PDT were transient chestdiscomfort and painful swallowing.

Pre-EMRdiagnosis Age Sex

Post-EMRDiagnosis

BE segmentlength

Post-PDT FollowUp

BE�HGD 69 yrs M Adenocarcinoma 3 cm 46 monthsBE�HGD 69 yrs M Adenocarcinoma 4 cm 13 monthsBE�HGD 68 yrs M Adenocarcinoma 4 cm 6 months

Conclusions: 1) The use of EMR in these BE�HGD patients with nodularor irregular mucosa resulted in histologic upstaging to adenocarcinoma andsubsequently higher laser light doses for PDT. 2) The use of porfimersodium PDT after EMR of superficial adenocarcinoma appears to be safeand effective for the complete elimination of Barrett’s dysplastic glandularmucosa. 3) EMR should be strongly considered for Barrett’s dysplasiapatients who are being evaluated for non-surgical management.

23

EFFECT OF INTRAESOPHAGEAL PERFUSION OFCHENODEOXYCHOLIC ACID IN PATIENTS WITH NON-EROSIVE GASTROESOPHAGEAL REFLUXAli Siddiqui, M.D., Sheila Rodriguez-Stanley, Ph.D.,Sattar Zubaidi, M.D., Susan Riley, Philip B. Miner, Jr., M.D., FACG*.OU Health Sciences Center, Oklahoma City, OK and OklahomaFoundation for Digestive Research, Oklahoma City, OK.

Purpose: Many patients with Non-Erosive Gastroesophageal Reflux Dis-ease (NERD) have relatively low esophageal acid exposure and respondsub-optimally to gastric acid suppression. In these patients, other constit-uents of gastric contents may induce esophageal symptoms. We havepreviously demonstrated that gastric contents can cause heartburn whengastric pH �4. (Aliment Pharm Ther 2000;14:129–134). The aim of thisstudy was to test esophageal sensitivity to chenodeoxycholic acid, a com-ponent of bile, administered as a provocative test.Methods: Patients with heartburn, absence of erosions, normal esophagealmotility, and % time esophageal pH �6% (NERD) were evaluated. Patientsunderwent a modified Bernstein acid infusion test and esophageal Barostatballoon distention. The Bernstein consisted of esophageal saline infusion (5min) followed by 0.1N HCI (10 min). Time and volume to 1st sensation/pain were recorded. Barostat balloon distention was performed using acylindrical high-compliance polyethylene balloon (5.5-cm length, 2.15-cmradius, 80-ml volume) placed 5 cm above the LES. 1st sensation/pain wererecorded. Sensitivity to 2 mM chenodeoxycholic acid (cheno) was assessedsimilar to the Bernstein acid infusion (10 ml/min). 2mM cheno was chosensince it is within the physiological range of bile salts in the stomach. Thevolume and time to 1st sensation and pain were measured and comparedwith the same end points during esophageal acid perfusion (Bernstein test).LS means were generated and one-tailed t-tests and regression analyseswere performed (P�0.05 level of significance).Results: 7 patients (2 Males, 5 Females; mean age 39.4 years) wereenrolled. Mean esophageal acid exposure was 3.4%. Mean volume ofballoon to pain was 28.6 ml (normal is � 50 ml). Time (min) to pain foracid and 2mM cheno were 5.6 and 6.4 min, respectively. Volume to painfor acid and cheno were 44.7 and 64.9 ml, respectively. 3 of the patientswere sensitive to saline infusion. Concordance between acid, cheno andBarostat sensitivity was not observed (P�0.05). Overall, patients withNERD were found to have variable responses to mechanical and chemicalstimuli.Conclusions: Esophageal perfusion of 2 mM chenodeoxycholic acid pro-vides an additional test of chemical sensitivity for esophageal sensitivitythat may be useful in assessing visceral pain, investigating pathophysiologyof NERD, and stratifying patients with NERD.

24

RABEPRAZOLE (RAB) PROVIDES GREATER DAY 1HEARTBURN (HB) RELIEF THAN OMEPRAZOLE (OME) INPATIENTS WITH ACUTE EROSIVE ESOPHAGITIS (EE) ANDMODERATE TO SEVERE SYMPTOMSStephen Fitzgerald, M.D.*, Richard Krause, M.D., FACG,Howard Schwartz, M.D., Gareth Lovett, M.D., Lian Mao, M.S.,Diane Hasner, Byron DeLemos, M.D. OMNI Healthcare, Inc.,Melbourne, FL; ClinSearch, Chattanooga, TN; Miami ResearchAssociates, Inc., Miami, FL; Eisai Inc., Teaneck, NJ and JanssenPharmaceutica Inc., Titusville, NJ.

S8 Abstracts AJG – Vol. 98, No. 9, Suppl., 2003

Purpose: Rapidity of HB relief is an important consideration in treatinggastroesophageal reflux disease (GERD). In previous studies, RAB dem-onstrated more rapid onset of effect on intragastric pH than OME. Thisstudy further investigated the impact of that difference on onset of symp-tom relief.Methods: This multicenter, randomized, double-blind trial enrolled 1129H pylori-negative patients aged 18–75 yr with �6-mo history of GERDsymptoms and �4 episodes of moderate/severe HB during an initial 7-dayplacebo run-in period. Patients also had either a documented history of EE(�5 yr) or acutely diagnosed EE (grade 2–4, modified Hetzel-Dent [HD]score) and received RAB 20 mg (n�559) or OME 20 mg (n�570) qd for7 days. Using an Interactive Voice Response System, patients recordeddaily severity of GERD symptoms on a 4-point Likert scale. The primaryendpoint was mean change from baseline in HB severity score for treatmentdays 1–3.Results: Overall, in the intent-to-treat (ITT) primary analysis population,mean change in HB severity score for RAB (�0.94) and OME (�0.93)showed a net improvement of 60% and 59%, respectively, over baseline(P�NS, RAB vs OME). Subgroup differences were noted between historicand acute EE: historic EE patients had higher prior proton pump inhibitorexposure (64% vs 34%), more females (50% vs 43%), and slightly greatertreatment response. With these differences, analysis focused on HB reliefin acute EE patients (Table). RAB was more effective for day 1 HB reliefespecially in patients with high-grade EE (HD grade 3–4): RAB, 31.5%versus OME, 16.5%. Both treatments were well tolerated.

Day 1 Complete HB Relief : Acute EE

Population

Net DifferenceRAB-OME

(% Patients)P-value

(�2 Test)

ITT (N�558) 7.2 0.057PP (N�528) 7.8 0.044ITT, High-Grade EE (N�189) 15.0 0.015

PP�per protocol.

Conclusions: Treatment with RAB and OME was effective for HB controlover time, but differential patient characteristics between historic and acuteEE groups may have reduced the ability to distinguish overall treatmenteffects. Complete HB relief with the first dose of RAB was greater inpatients with acutely confirmed EE and nearly twice that of OME inhigh-grade EE patients. Research supported by Eisai Inc., Teaneck, NJ, andJanssen Pharmaceutica Inc., Titusville, NJ.

25

ESOPHAGEAL ACID PERFUSION IN CONDITIONEDATHLETES WHILE RUNNING DECREASES PERFORMANCEIN SUBJECTS WITHOUT A HISTORY OF HEARTBURN, BUTNOT IN RUNNERS WITH HEARTBURN HISTORYMandy R. Graves, B.S., Philip B. Miner, Jr., M.D., FACG*,Debra Bemben, Ph.D., Aaron M. Smathers, Sattar M. Zubaidi, M.D.Oklahoma Foundation for Digestive Research, Oklahoma City, OK andUniversity of Oklahoma, Norman, OK.

Purpose: We have recently demonstrated increased acid reflux into theesophagus of conditioned athletes during exercise. Extraintestinal manifes-tations of GERD include asthma and laryngitis suggesting acid reflux intothe esophagus may alter athletic performance. The aims of this study were1) to compare ‘exercise to exhaustion’ in conditioned athletes with andwithout a history or heartburn while the esophagus is perfused with saline(sham) or 0.1 N HCl (Bernstein) and 2) to assess the effect of acid perfusionon pulmonary function after exercise.Methods: 24 conditioned runners (13 no heartburn; 11 heartburn history)with no significant medical history underwent a history and physical

examination, screening laboratory tests and a treadmill test for cardiovas-cular safety. Subjects were randomly assigned to receive either shaminfusion or acid infusion and subsequently crossed-over. Pulmonary func-tion tests (PFTs) were performed, the esophagus intubated with smallcaliber, flexible catheter. The infusion was begun with an IV infusion pumpat 8 ml/minute. The subject was exercised to exhaustion using a steady statesubmaximal (85% of exercise capacity) endurance protocol. The time toexhaustion was recorded and PFTs were measured immediately after ex-ercise.Results: The mean time to exhaustion for subjects with heartburn wasnearly identical with sham and Bernstein infusions (19.58 vrs 19.83 min-utes; p�NS). Esophageal acid perfusion in normal subjects decreased thetime to exhaustion by more than 10% from 23.13 minutes (sham) to 20.66minutes (Bernstein) (p�0.041). PFTs were not altered by acid infusion.Conclusions: Acid perfusion of the esophagus significantly decreasedexercise ability in subjects without heartburn symptoms and did not influ-ence exercise tolerance in subjects with heartburn. An acid induced changein pulmonary function does NOT explain the change in exercise tolerance.

26

HEARTBURN FREQUENCY CAN PREDICT THE PRESENCEOF PATHOLOGIC ESOPHAGEAL REFLUX IN GERDPATIENTS WITHOUT ESOPHAGITIS DURING TREATMENTWITH RABEPRAZOLEJerry D. Gardner, Sheldon Sloan, Malcolm Robinson,Philip B. Miner, Jr.*. Science for Organizations, Inc., Chatham, NJ;Janssen Pharmaceutica, Inc, Titusville, NJ and Oklahoma Foundationfor Digestive Research, Oklahoma City, OK.

Purpose: Although proton pump inhibitors are known to reduce esopha-geal acid reflux and the accompanying heartburn in patients with GERDwithout esophagitis, no quantitative relationships between heartburn andesophageal reflux have been established. In the present study, we examinedpossible relationships between heartburn frequency and esophageal acidreflux in endoscopy-negative GERD subjects treated with a rabeprazole.Methods: Subjects (N�25) had GERD without esophagitis (endoscopy 0or 1 on Hetzel-Dent scale) and experienced heartburn at least 5 timesduring a 7-day run-in period. Esophageal pH was recorded for 24 hoursbefore and after 4 weeks treatment with rabeprazole 10mg/day (N�9),rabeprazole 20mg/day (N�8) or placebo (N�8) in randomized, double-blind fashion. The number of episodes during the last 7 days of treatmentwere used for analysis.Results: Using receiver operating characteristic analysis, the optimal valuefor heartburn frequency that distinguished between subjects with andwithout pathologic esophageal reflux was 12 episodes/week. This value hada sensitivity of 89% and a specificity of 47% for pathologic esophagealreflux. We also calculated the probability of pathologic esophageal refluxfor each value of heartburn frequency and found that the probability ofpathologic esophageal reflux could be related to gastric pH by an equationwith a single exponential function. This equation fit the data with R2 of0.94. With heartburn frequency of 3 episodes/week or less, the calculatedprobability of pathologic reflux was less than 5%. There was close agree-ment between predicted and observed percentages of subjects with patho-logic esophageal reflux with the different active treatments. For example,with 20mg rabeprazole, the percentage of subjects with no pathologicreflux predicted on the basis of heartburn severity was 90% [81,99] (mean,95% CI) and the observed percentage was 89% [68,110].Conclusions: The present results indicate that in endoscopy-negativeGERD subjects treated with rabeprazole, heartburn frequency can predictthe likelihood of pathologic esophageal reflux over a 24-hour period. Suchan assessment can be used to adjust the dose of a rabeprazole in patientswith persistent symptoms.

S9AJG – September, Suppl., 2003 Abstracts