109

Can two upper endoscopies negative for dysplasia eliminate the needfor futiure surveillance in patients with Barrett’s esophagus?Prateek Sharma*, Allan Weston, Gary Falk, Mark Johnston, Dean Rekerand Richard Sampliner. 1Gastroenterology, VA Medical Center &University of Kansas, Kansas City, MO; 2Cleveland Clinic Foundation,Cleveland, OH; 3Bethesda Naval Medical Center, Bethesda, MD; and4Southern Arizona VA HealthCare System, Tucson, AZ.

Purpose: Surveillance endoscopy in patients with Barrett’s esophagus(BE) is undertaken for the early detection of dysplasia and cancer; however,the time interval of progression to high-grade dysplasia (HGD) or cancer isnot known and surveillance intervals not clearly defined. The aim was toevaluate if patients with at least 2 upper endoscopies (EGD’s) negative fordysplasia progress to HGD and/or cancer.Methods: This study is a multi-center clinical and endoscopic outcomesproject involving a single large database of patients with BE; currentlyincluding Kansas City VAMC, Bethesda Naval Medical Center, ClevelandClinic and Southern Arizona VA Health Care System. The initial goal ofthis project was to define the incidence and prevalence of HGD and cancerin patients with BE. Data from each of the participating centers weremerged and generated into the main study database using Microsoft Ac-cess. Preliminary data from each center included patient age, gender,ethnicity, endoscopy dates with dysplasia grades and Barrett’s length. Astandardized definition of BE columnar lined distal esophagus of any lengthwith intestinal metaplasia (IM) was used as inclusion criteria. Patients withat least 12 months of follow up were included in the long term follow upgroup. Cancers and HGD occuring within 12 months of the index EGD/biopsy (bx) were regarded as prevalence cases.Results: 1376 patients met study criteria and had at least EGD/bx revealingBE. Of these 618 patients (95% Caucasians, 14% Females) have beenfollowed for 2546 patient years; mean f/u 4.12 years (range: 1–22.5 years).Twelve patients developed cancer during f/u (incidence 0.5% per year) and22 pts developed HGD (incidence 0.9% per year)-a total of 34 incidenceHGD/cancers. Of these 34 patients developing either HGD and/or cancer,18 patients (53%) had at least 2 initial consecutive EGD/bx revealing IMwithout dysplasia over a mean time period of 2 yrs (range: 1 month-3yrs).After the 2nd consecutive EGD/bx revealing IM without dysplasia, HGD/cancer was detected after a mean of 3 yrs (range: 3 months-7.8 yrs).Conclusions: In this large cohort of patients with BE, at least half thepatients who developed HGD and/or cancer had 2 consecutive initialEGD/bx revealing IM without dysplasia. These results support that even inpatients with 2 EGD/bx with IM without dysplasia, continued surveillanceis indicated. Factors other than a negative initial dysplasia grade need to beevaluated for risk stratification for HGD/cancer developement. (Funded byADHF/ASGE)

110

Control of esophageal acid exposure in patients with Barrett’sesophagus on rabeprazolePrateek Sharma*, Allan Weston, Shannon Keeton, Lisa Camargo andRichard Sampliner. 1Gastroenterology, VA Medical Center &University of Kansas, Kansas City, MO; and 2Gastroenterology,Southern Arizona VA HealthCare System, Tucson, AZ.

Purpose: Patients with Barrett’s esophagus (BE) represent the severe endof the GERD spectrum with abnormal esophageal acid exposure and loweresophageal sphincter (LES) dysfunction. Some studies have indicated thatGERD patients with H pylori infection may have “improved” gastric acidcontrol on proton pump inhibitor (PPI) therapy. We evaluated the results ofambulatory 24-hr pH monitoring in patients with BE treated with rabepra-zole 20 mg BID and their H pylori status.Methods: Patients with BE with either no dysplasia or low grade dysplasiaand a segment length of 2–6 cms were invited to participate in an endo-scopic ablation study while on rabeprazole 20 mg BID. BE was diagnosedby the presence of esophageal columnar mucosa with intestinal metaplasia

on biopsy. After at least 7 days of rabeprazole therapy, patients underwentambulatory 24-hr pH monitoring (on PPI). H pylori status was determinedby obtaining 4 biopsies from the gastric cardia and 4–8 biopsies from thegastric corpus and antrum.Results: Twenty four patients have been enrolled in the study; 22 males/2females, mean age 61.9 years (range: 42–84). Six of 24 patients (25%) hadan abnormal 24-hr pH result. The results of ambulatory 24-hr pH moni-toring are as follows: Mean total pH �4: 4.7% (range: 0–29.9%); meanupright pH �4: 3.8% (range: 0–47.9%); mean supine pH �4: 5.3% (range:0–40%). Only 2 patients (8.3%) were H pylori positive, both had normal24 hr pH results.Conclusions: Seventy five percent of patients with BE had “normalization”of esophageal acid exposure on BID rabeprazole therapy. H pylori wasdetected in only 8% of patients with BE in this study group-only a smallpercentage of pateints with BE have H pylori, indicating that H pyloriprobably plays a minor role in “acid control” in BE patients. (Funded byJanssen/Eisai)

111

Detection of cardia intestinal metaplasia: Does the biopsy numberand location matter?Prateek Sharma*, Margarita Topalovski, Shannon Keeton and AllanWeston. 1Gastroenterology & Pathology, VA Medical Center & Univ ofKansas, Kansas City, MO.

Purpose: Presence of intestinal metaplasia in the gastric cardia (CIM) hasbeen reported in 2–24% of patients undergoing EGD and is a topic ofinterest given the rising incidence of cancer in this location. Our aim wasto determine the prevalence of CIM in biopsies obtained from two seperatelocations within the cardia.Methods: Patients presenting to the endoscopy unit for EGD for anysymptoms were invited to participate in the study. The biopsy protocol was:8 from the gastic cardia-4 from upper cardia (forceps across SCJ-partsquamous/part columnar), 4 from lower cardia (within 1 cm of uppercardia), 4 body and 4 antrum. All cardia biopsies were stained with H&Eand alcian blue at pH 2.5 for the presence of IM; body/antrum biopsiesstained with Steiner stain for H. pylori (Hp) detection. Patients testingnegative for Hp on biopsies, underwent a serology test (Flexsure). A singlepathologist evaluated all specimens.Results: Sixty-five patients have been evaluated by this protocol; medianage 54 yrs (range: 34–81), 63 males, 53 Caucasians and 12 AfricanAmericans. 80% of upper cardia biopsies had both cardiac�squamoustissue. Detection of CIM was as follows:

Biopsy Location No. of Pts With CIM

Upper cardia only 7Both upper & lower cardia 5Lower cardia only 7

Thus, the overall prevalence of CIM was 29% (19 patients); with CIMdetected in 12 patients (18%) in the upper cardia and in 12 patients (18%)in the lower cardia. Hp was detected in 58% of the patients with CIM byeither histology or serology; the addition of serology detected 8 additionalHp positive patients.Conclusions: Prevalence of CIM in this study was similar at each locationin the cardia (18%, 4 biopsies); however, if both locations were considered,prevalence increased to 29% (8 biopsies). Thus CIM prevalence may varydepending on the number of biopsies and areas biopsied. Use of additionaltesting detects more patients who are Hp positive and should be performedin determining the association of CIM with Hp. Future endoscopic studiesof the gastric cardia should specify the location/number of biopsies and thetests used to diagnose Hp. (Funded by AstraZeneca)

S36 Abstracts AJG – Vol. 96, No. 9, Suppl., 2001