A884 AGA ABSTRACTS GASTROENTEROLOGY Vol. US, No.4

4890THE EFFECTS OF LONG-TERM OMEPRAZOLE THERAPY ONGASTRIN, PEPSINOGEN I, AND CHROMOGRANIN A CONCEN­TRATIONS ARE RELATED TO CYP2C19 POLYMORPIDSM.Mohamed Eg Sagar, Leif Bertilsson, Mats Stridsberg, Sven Mard, AnnKjellin, Rein Seensalu, Karolinska Inst, Stockholm, Sweden; Uppsala UnivHosp, Uppsala, Sweden; Linkoping Univ, Linkoping, Sweden.

Background: Omeprazole is metabolized by CYP2CI9 in the liver. Previ­ously we have shown a more marked effect after 8 days omeprazoletreatment on plasma gastrin and intragastric acidity in healthy subjects andpatients with one functional CYP2C19 allele (wtlmut) compared to thosewith two functional alleles (wtlwt). Aim: To investigate the effect oflong-term therapy of omeprazole on serum gastrin, pepsinogen I, andplasma chromogranin A (CgA) in patients with gastroesophageal refluxdisease (GERD) with different CYP2C19 genotypes. Methods: Serumgastrin, pepsinogen I and plasma CgA concentrations were measured in 71patients with gastroesophageal reflux disease on omeprazole treatment (20mg od) for more than one year (median 32 months). Genotyping for defectalleles was performed by PCR amplification. Serum gastrin and CgA wereassessed with specific radioimmunoassays. Pepsinogen I and H. pyloriserology were determined by validated ELISA methods. Results: Nineteenout of the 71 patients were heterozygous for CYP2C19ml (wtlmut), and 52were homozygous for the wild-type alleles in both exon 4 and 5 and(wtlwt). Fourteen patients were H. pylori positive (13 wtlwt, and onewtlmut). Significant differences in serum gastrin, pepsinogen I and CgAlevels were observed between the two groups as shown in the table below.Conclusion: The marked increase in gastrin and CgA concentrations in thewtlmut indicate that a more effective acid inhibition in these patientsaffects the endocrine cells in the antral and oxyntic mucosa to a greaterextent than in patients with a normal omeprazole metabolism (wtlwt). Thedecrease in pepsinogen I level, in the same group of patients, may indicatesome degree of gastric atrophy related to omeprazole treatment.

4889BLOCKAGE OF 5HT3 PATHWAYS REDUCES GASTROESOPH­AGEAL REFLUX (GER) INDUCED BY INTRADUODENALCHOCOLATE.B. H. Kim, Han-chun Lien, Wei Ming Sun, Mark McDermott, ChungOwyang, Univ of Michigan, Ann Arbor, MI.

Certain foods such as chocolate are known to induce GER symptoms, i.e,heartburn. The mechanism responsible is unknown. Serotonin (5HT) re­leased from the intestinal enterochromaffin (EC) cells appears to be acommon mediator for many luminal stimuli that activate the vagal path­way. We hypothesize that chocolate releases 5HT from intestinal EC cells.This in tum activates 5HT3 receptors on vagal mucosal afferent fiberswhich induce relaxation of LES and reduce esophageal peristalsis via avago vagal pathway. To test this hypothesis we performed esophagealmanometry, esophageal acid clearance studies and intraesophageal pHmonitoring in six healthy volunteers. On separate days, each subjectreceived intraduodenal infusion of chocolate (1.8 kcallmin, 3 ml/min) or3% NaCI (equiosmotic to chocolate) with or without i.v, granisetron (10JLg/kg, a 5HT3 antagonist). LES pressure was measured by esophagealmanometry and esophageal acid clearance was evaluated by measuring theduration for pH to return to 4.0 after intraesophageal infusion of 15 ml of0.1 N HCI. GER was recorded by a pH probe placed 5 cm above the LES.Results: Chocolate significantly increased reflux events and acid exposuretime when compared to 3% NaCl. Granisetron reduced acid exposureinduced by chocolate. Compared to placebo (saline). granisetron alsodecreased acid clearance time during chocolate infusion (Table). Further­more, chocolate reduced LES pressure in 4 of 6 subjects which wasreversed by granisetron. In conclusion, we demonstrated that chocolateinduces GER symptoms by reducing LES pressure. increasing reflux eventsand acid exposure duration. It also increases acid clearance time. All theseevents could be antagonized by granisetron. Hence blockade of 5HT3

pathways may be a novel approach to the treatment of GER disease.

'p<0,05, vs 3% NaGI, 'p<0,05, vs chocolate

Reflux event/30 minAcid exposure time (%)Acid clearance time (min)

Chocolate (id)

6.5 ± 19'23.9 ± 3,3'24 ±6.5'

Chocolate Od)Granlsetron (Iv)

3,3 ± 0.9'11.2± 3,9'7.0±2.5'

3% MaCI (Id)Control

0,8 ± 0.75.9 ± 0.4115± 6.9

4891

ROLE OF BILE ACID MALABSORPTION AND CLINICAL FEA­TURES IN LYMPHOCYTIC COLITIS. A COMPARISON WITHCOLLAGENOUS COLITIS.Kjell-Ame Ung, Anders Kilander, Roger Willen, Hasse Abrahamsson,Internal Medicine, Goteborg, Sweden; Pathology, Goteborg, Sweden; Univof Goteborg, Goteborg, Sweden.

In a recent study we found a high prevalence of bile acid malabsorption anda high symptomatic response rate to bile acid binder treatment in patientswith collagenous colitis (CC). The knowledge of the role of bile acids inlymphocytic colitis (LC) is limited. If LC and CC are parts of a spectrumof the same disease or not is a matter of debate. The purpose of this studywas to determine the role of bile acid malabsorption in LC and to evaluatethe relationship between LC and CC. Methods: Patients with chronicdiarrhoea fulfilling the criteria for LC (n =23) completed a diagnosticprogramme including the 75SeHCAT test. Symptoms were registered in aquestionnaire during one week. Prevalence of bile acid malabsorptiondefined as 75SeHCAT :s 10 % day 7, response to bile acid binders andclinical data were compared with previosly obtained data in 27 patientswith CC. The 75SeHCAT values in LC, CC and 29 healthy controls werecompared. Results: Bile acid malabsorption was significantly less commonin LC than in CC (2/23 V.s. 12/27) (p < 0.01). The 75SeHCAT values inLC (median (range): 24 % (1.7 - 53) were higher than in CC (median(range): 12 % (0.5 - 41» (p < 0.008) but lower than in the control group(median (range): 38 % (8 - 91» (p < 0.006). Fourty-six % with LCresponded to bile acid binders v.s. 78 % with CC (p = 0.072). One patientwho was recolonoscoped after 10 months had converted from LC to CC.There was a trend to a higher prevalence of smokers in CC than in LC(14/27 v.s. 6/23) (p = 0.086). There were no differences in age, gender,occurrence of immunologic diseases or symptoms including stool fre­quency and consistency, pain, distension and flatulence between LC andCc. Conclusion: Bile acid malabsorption is less common in LC than in Cc.The lower 75SeHCAT values in LC compared to the control group suggesta possible pathophysiological role of bile acids in LC. Bile acid bindersshould be considered as a treatment option in LC. The conversion of onepatient from LC to CC and no clinical differences except for less disturbedbile acid excretion suggest that LC and CC might be different spectra of thesame disease.

4892EFFECT OF CHOLECYSTECTOMY ON BILE ACID METABO­LISM AND SERUM LIPOPROTEIN PROFILES.Cyrus A. Moussavian, Klaus Parhofer, Dieter Juengst, Med Dept II, UnivMunich, Munich, Germany.

Bile acid malabsorption (BAM) has been supposed to play an importantpathogenetic role in postcholecystectomy diarrhea (PCD). It remains un­clear, however, whether BAM also occurs in asymptomatic patients afterCHE. To evaluate the effect of CHE on bowel habits, bile acid metabolism,and serum lipoprotein profiles, we performed a prospective case-seriesstudy in patients undergoing CHE. The loss of bile acids from the entero­hepatic circulation was determined by the use of serum concentrations of70'-hydroxy-4-cholesten-3-one (70'-HC). Patients and Methods: Bowelhabits and biochemical parameters including serum levels of 70'-HC weredetermined in 48 patients (mean age 58.6:'::15.2 years, females/males32/16) before, at 4 weeks, and at 12 weeks after elective CHE. 70'-HC wasextracted from serum with ODS-bonded silica at 64 °C and determined byHPLC and UV-detection. Statistics were performed with the McNemar¥-test for discrete values and Student's paired t-test for continuous values;P-values <0.05 were considered statistically significant. Results: At 4 and12 weeks after CHE, the number of patients reporting more than 1 bowelmovement per day increased from 12/48 (25 %) to 24/48 (50 %, P<O.OOI)and 21/48 (43 %, P<0.005), respectively, while the number of patientsreporting formed stools decreased from 47/48 (98 %) to 28/48 (58 %;P<O.OOI) and 33/48 (69 %, P<O.OOI), respectively. There were no patientswith diarrhea evolving after CHE. Serum levels of 70'-HC increased from23.4± 15.9 ng/ml to 48.9±30.9 ng/ml (P<O.OOI), and 55.2±34.8 ng/ml(P<O.OOI), respectively. Serum levels of HDL-cholesterol increased from46.6±17.6 mg/dl to 48.5±15.3 mg/dl (P<0.05) and 51.2±16.1 mg/dl(P<O.OI), respectively, while serum levels of cholesterol, LDL-choles­terol, and triglycerides remained unchanged after CHE. Conclusion: CHEresults in an increased loss of bile acids from the intestine which isreflected by a compensatory increase of bile acid synthesis and changes instool frequency and stool consistency. The long-term effect of thesechanges, which seem to be associated with a slight increase of HDL­cholesterol in serum, remain to be elucidated.

wt/wt wt/mutn n p

Gastrin (pM) 52 14.7±0,9 19 52.1±7.6 0,0001CgA(nM) 52 2.5±0.1 19 7,3±1.3 0.0001Pepsinogen I btg/I) 49 193±12 19 147±19 0,04

Mean±SE, Mann Whitny U-test.