SUMMARY INTRODUCTION. Aliment Pharmacol Ther 2000; 14: 625±634. Accepted for publication 20 January 2000

Transcription

1 Aliment Pharmacol Ther 2000; 14: 625±634. Cure of Helicobacter pylori infection in atrophic body gastritis patients does not improve mucosal atrophy but reduces hypergastrinemia and its related effects on body ECL-cell hyperplasia B. ANNIBALE, M. ROSARIA APRILE, G. D'AMBRA, P. CARUANA, C. BORDI & G. DELLE FAVE Gastroenterology Unit, University `La Sapienza' Rome, Department of Pathology University of Parma, Italy Accepted for publication 20 January 2000 SUMMARY Background: The effects of H. pylori eradication on atrophic body gastritis are controversial. Aim: To investigate the effect of triple therapy on atrophic body gastritis in H. pylori-positive patients and its effect on morpho-functional gastric parameters. Methods: Thirty- ve consecutive atrophic body gastritis patients with histological/serological evidence of H. pylori infection were treated. Before and 6 and 12 months after H. pylori eradication the patients were evaluated for fasting gastrinemia and pepsinogen I, basal and peak acid output, and detailed histological assessment including the ECL cell proliferative patterns. Results: Six months after treatment, 25 out of 32 patients were cured (78%). Cure of infection was associated with improvement in both basal (basal acid output mean mmol/h vs mmol/h, P < 0.005) and stimulated acid secretion (peak acid output mean mmol/h vs mmol/h, P ˆ ) as well as with reduction in hypergastrinemia (mean gastrin levels pg/ml vs pg/ml; P < 0.005). In contrast, the eradication had no effect on body corporal atrophy and intestinal metaplasia, or pepsinogen I levels (mean ng/ml vs ng/ml, N.S.). These results were con rmed at 12 months after eradication. A statistical inverse correlation was obtained (r ˆ ±0.3635, P < 0.05) between the corporal chronic in ltrate score and peak acid output values. A total of 53% of atrophic body gastritis patients showed a regression in severity of body ECL cell hyperplastic change. Conclusion: Cure of H. pylori infection in patients with atrophic gastritis reverses some adverse effects on gastric function and ECL cell hyperplasia. H. pylori infection may be cured in atrophic body gastritis patients with partial reversion of its negative consequences on acid secretion and body ECL cell hyperplasia. INTRODUCTION Atrophic body gastritis is associated with gastric autoimmunity, which eventually leads to mucosal atrophy and, in an unknown number of cases, to de ciency of intrinsic factor and vitamin B 12, eventually resulting in pernicious anaemia (PA) 1, 2. The way in which this type of gastritis initiates and evolves is so far Correspondence to: Dr B. Annibale, Cattedra di Gastroenterologia I, Clinica Medica II Policlinico Umberto I, Rome, Italy. annibale@uniroma1.it unknown, but a deep plasma cell and lymphocytic in ammation presumably leads to the loss of oxyntic glands, with a consequent total disappearance of parietal and chief cells and thus, of the production of acid and intrinsic factor. 3 Due to the lack of acid inhibitory feed-back this event induces a sustained release of circulating gastrin. 4 It is now apparent that Helicobacter pylori is also involved in the induction of atrophic body gastritis. 5, 6 In fact, some studies have recently provided evidence that long-standing uncured H. pylori-antral gastritis could have, as a sequel, Ó 2000 Blackwell Science Ltd 625

2 626 B. ANNIBALE et al. H. pylori-related atrophy of the body mucosa. 7, 8 We have previously shown a 26.8% prevalence of H. pylori infection in a group of newly diagnosed patients with histologically proven atrophic body gastritis and with functional evidence of atrophy identi ed by means of a screening study in patients with unexplained anaemia or long-standing history of dyspepsia. 9 Furthermore, it has been reported by us and others that cure of H. pylori infection in atrophic body gastritis patients is associated with partial reversion of achlorhydria and amelioration of the histological corporal in ammation. 7, 9, 10 However, these results were obtained in a small number of patients or in hyposecretory patients and discrepancies still existed in the ability of eradication to 7, 9, 10 reverse the corporal atrophy. The aim of this study was to investigate in a consecutive population of H. pylori-infected patients with atrophic body gastritis the ability of triple therapy to cure H. pylori infection and to verify whether H. pylori eradication may modify the gastric functional data (acid secretion and gastrinemia) and the histological patterns (atrophy, intestinal metaplasia and enterochromaf nlike cells hyperplasia) related to the corporal mucosal atrophy. 1MATERIALS AND METHODS Patients On the basis of a screening system for atrophic body gastritis as previously reported 122 consecutive new out-patients with atrophic body gastritis were diagnosed. 9 Brie y, all patients, referred to us from the Haematological Department for macrocytic or microcytic anaemia (n ˆ 99) or from our Department for unexplained long-standing dyspepsia (n ˆ 23), were screened for fasting blood gastrin levels; if these values were outside the normal range (0±40 pg/ml) they were further submitted to gastric secretory tests and gastroscopy with multiple biopsies in order to ascertain the presence of atrophic body gastritis. Patients taking antisecretory drugs or with previous gastro-oesophageal surgery were excluded. Among the 122 atrophic body gastritis patients, evidence of H. pylori infection was documented in 35 patients (28.6%). Design All 35 atrophic body gastritis patients had a detailed initial assessment including measurement of H. pylori immunoglobuline (Ig)G serology, parietal cell and intrinsic factor antibodies and fasting gastrin, upper gastrointestinal endoscopy and biopsies from the 2antrum and body of stomach for histology and culture. They also had basal and peak acid output to pentagastrin. These tests are described in detail later. After the initial assessment all 35 patients were treated with H. pylori eradication therapy and re-examined 6 months after the start of treatment. Furthermore, 15 atrophic body gastritis patients successfully cured of infection agreed to be re-examined 6 months later, after documented eradication (12 months after the start of treatment), with acid secretory tests and gastroscopy with biopsies from the antrum and body of stomach for histology. All patients gave informed consent to the study. The study was approved by the local ethical committee. Criteria for the diagnosis of atrophic body gastritis and pernicious anaemia. The diagnosis of atrophic body gastritis was based on the pentagastrin-resistant achlorhydria or hypochlorhydria (peak acid output lower than 13 mmol/h) and histological con rmation of atrophy (see histological procedure section). 9, 11 The diagnosis of pernicious anaemia was based on the following criteria: pentagastrin-resistant achlorhydria; macrocytic anaemia; presence of Intrinsic Factor antibodies; and response to vitamin B 12 therapy, as 9, 12 described. Histological procedures. Three biopsy samples from the antrum (smaller and greater curve, anterior or posterior walls) and four from the mid-body along the greater curve were taken with standard biopsy forceps. 13 The specimens were immediately xed in Bouin's solution for 4±8 h at room temperature, rinsed in 0.1 M phosphate-buffered saline, ph 7.4 and routinely processed to wax. Serial paraf n sections (5 lm) were stained with haematoxylin-eosin for conventional histological determination and with Giemsa stain for H. pylori evaluation. Assessment of the degree of gastritis was performed according to the updated Sydney System. 14 To each graded variable the following scores were assigned: 0 for absence, and 1, 2, 3 for mild, moderate or severe presence, respectively. Atrophy of the fundic mucosa was de ned as the focal or complete replacement of oxyntic glands by metaplastic pyloric or intestinal glands. Atrophy of the antral mucosa was de ned as

3 CURE OF H. PYLORI IN ATROPHIC BODY GASTRITIS 627 focal or complete replacement of antral glands by intestinal metaplastic epithelium. The biopsies were examined independently by two histopathologists (CB, PC), who were not aware of the clinical data of the patients. In the case of disagreement, the relevant biopsies were re-examined simultaneously by both histopathologists until an agreement was reached. Immunocytochemistry of ECL proliferative pattern. Serial 5-lm thick sections of gastric body mucosa perpendicular to the mucosal surface were stained for immunostaining of endocrine cells using monoclonal antibody against chromogranin A (CgA; clone LK2H10, Biogenex Laboratory, San Ramon, CA) followed by the avidin±biotin complex procedure (Dako, LSAB2 kit, Dakopatts, Glostrup, Denmark). Biotinylated goat antirabbit IgG (code BA1000, Vector Laboratories, Burlingame, CA; dilution, 1:200) were then used as a secondary antibody, followed by the avidin±biotin complex peroxidase technique (Vectastain ABC Kit, Vector Laboratories, Burlingame, CA). Diaminobenzidine tetrahydrochloride was used as a chromogen substrate and nuclear counter-staining with haematoxylin was performed. The ECL cell changes were evaluated in non-intestinalized areas of gastric body mucosa. The patterns of endocrine cell growth were identi ed according to our recent revised version of the classi cation proposed by Solcia et al. and independently evaluated by two experienced pathologists (CB, PC) 11, 15 who were unaware of the patients' clinical data. H. pylori status. Patients were de ned having H. pylori infection by a single positive result in either histology, 3serology or culture. Eradication of H. pylori infection. All 35 H. pylori-positive patients with atrophic body gastritis agreed to be treated with triple therapy: 120 mg q.d.s. of dicitrate bismuthate for 4 weeks, plus amoxicilline 1 g t.d.s. and metronidazole 250 mg t.d.s. during the rst 2 weeks of therapy. Eradication of infection was assessed 6 months after the beginning of treatment. To this purpose a systematic sampling of the gastric mucosa including four biopsies from antrum and seven from fundus/corpus, as recommended by Genta was performed. 16 Moreover, tissue culture from the antrum and corpus was performed. The negativity of two tests associated with at least a 50% decrease in initial titre of H. pylori IgG were accepted as criteria for infection cure. Investigative tests Gastric acid secretion. Basal acid output and pentagastrin-stimulated (peak acid output) (6 lg subcutaneously Peptavlon, Zeneca, England) gastric acid secretions were determined in all patients, as previously described. 17 The reference values of our laboratory were, respectively: basal acid output, median 4.1 mmol/ h (range 1.2±11.2 mmol/h); and peak acid output, median 31.1 mmol/h (range 13±44.4 mmol/h). 18 Microbiology. Culture was performed on blood agar (7% of de brinated horse blood). After incubation in 10% CO 2 at 37 C for 5 days, bacterial growth was examined by Gram staining to reveal the typical spiral organisms; urease, catalase and oxidase tests were performed to identify the organism as H. pylori as previously described. 9 Plasma studies. Fasting gastrin levels were evaluated in plasma by means of a speci c radioimmunoassay using antibody no (Professor J. F. Rehfeld), as described earlier. 17, 18 Normal reference values: 0±40 pg/ml. H. pylori IgG antibodies were determined using an ELISA commercial kit (G.A.P. test IgG, Biorad, Milan, Italy) as previously reported. 18 The positive cut-off for H. pylori IgG was taken as 13 IU/L. Pepsinogen I levels were measured using commercial kit RIA (Pepsik, Sorin, Saluggia, Italy) as reported earlier. 18 The normal reference values for our laboratory were 20±80 ng/ml. Parietal cell antibodies were determined on serum using a solid phase immunosorbent assay commercial Kit (AUTOSTAT, Cogent Diagnostic Ltd, Edinburgh, England) with a variation coef cient of 14.3%. 9 Intrinsic factor antibodies (IFA) were determined using a commercial Kit (IFbAb, Diagnostic Products Corporation, Los Angeles, CA) solid phase radioassay for the detection of blocking antibody in serum. 9 Statistical evaluation Data were expressed as mean S.E.M. and/or median (range) and evaluated by appropriate statistical tests (t-test for paired data or Wilcoxon test). Subgroups

4 628 B. ANNIBALE et al. (percentages) in the two populations were compared by means of Fisher's exact test. A P-value of less than 0.05 was considered statistically signi cant. RESULTS Table 1 details the anagraphical and biochemical data of the 35 patients with histologically proven atrophic body gastritis and H. pylori infection: The median age at diagnosis was 49 years (range 22±73). Parietal cell antibodies were present in 51.4% of patients but only two patients had pernicious anaemia as shown by intrinsic factor antibodies with associated macrocytic anaemia. The great majority of patients (28 out of 35; 80%) had histological detection of H. pylori in body mucosa (mean colonization score for body and antrum was , and , respectively). Two patients had only H. pylori culture positive and ve patients had only elevated title of H. pylori IgG. Furthermore only four atrophic body gastritis patients had severe body atrophic changes (score 3) (Table 1). Median fasting gastrin levels was 280 pg/ml (range Table 1. Anagraphical and biochemical characteristics of atrophic body gastritis H. pylori-positive patients Case Age/ sex Gastrin pg/ml 16 Ab PCà 17 IFAà Body H. pylori* Body histology Body culture Ab H. pylori IgG U/L 1 66/F 135 ) ) /F 88 + ) /f 50 ) ) /F 70 ) ) 3 2 ) /M ) /F 110 ) ) /M ) 1 2 ) /F /M 70 ) ) /F ) 1 2 ) /F 230 ) ) 2 1 nd /F 300 ) ) 1 2 ) /F ) ) /F 100 ) ) 3 2 nd /F ) /M 700 ) ) 1 3 nd /F 220 ) ) 1 2 ) F ) /f ) /f ) 2 1 ) /f 236 ) ) 1 1 ) /F ) 2 2 ) /F 650 ) ) 2 2 ) /F 93 ) ) ) 1 ) /F 150 ) ) ) 2 ) /F ) ) /F ) /F 1650 nd ) 1 1 nd /F ) ) 2 nd /F ) ) 1 nd /M 77 ) ) 1 3 ) /F ) 1 2 ) /F 500 ) ) 2 2 nd /F 262 ) ) ) 2 ) /F ) 1 2 ) 13 * Sydney Score of atrophy. Sydney Score of H. pylori colonization. à Normal value = negative. nd, not determined.

5 CURE OF H. PYLORI IN ATROPHIC BODY GASTRITIS ±1650 pg/ml) and median Pepsinogen I levels was 16 ng/ml (range 15±41 pg/ml). Eighteen patients were achloridric, while seventeen patients had a persistence of low acid secretion: median peak acid output 2.2 mmol/h (range 0±11.6 mmol/h). Cure of H. pylori infection Effect of treatment. Thirty- ve atrophic body gastritis H. pylori-positive patients were treated with triple therapy (see material and methods). Three patients were lost at the follow-up because they refused endoscopy. No patients stopped the treatment for side-effects. However, mild side-effects (diarrhoea, metallic taste, abdominal pain) were present in 23 (65.7%) patients whereas only two (5.7%) patients had moderate sideeffects. Gastric histology at 6 months. Out of 32 patients, 25 were successfully cured (78.1%) with no further evidence of active gastritis in the corpus and the antrum 6 months after the treatment. In fact, the gastric corpus mucosa (Figure 1A) showed a complete resolution of acute in ammation and a signi cant reduction in chronic in ammation with respect to the 4pre-treatment (Mean scores: vs ; P < 0.001). Atrophy and intestinal metaplasia were not affected. Their pre vs. post-eradication 5mean scores were: vs and vs , respectively. Antral mucosa (Figure 1B) revealed the same response observed on the body mucosa. Acute in ammation disappeared and chronic in ammation was signi cantly reduced with respect to the pre-treatment whereas antral atrophy and metaplasia were not modi ed (mean 6scores: vs , N.S., and vs , N.S., respectively). In the seven atrophic body gastritis patients who were notcured, none of the graded variables of corporal and antral Sydney score were modi ed (data not shown). Serum assays at 6 months. Fasting gastrin in the 25 cured patients, compared to the values obtained before therapy, showed a signi cant decrease (mean gastrin levels: pg/ml vs pg/ml; P < 0.005). In contrast, Pepsinogen I levels were not modi ed by the treatment (mean ng/ml vs ng/ml; N.S.) according to the unchanged atrophy score. In the seven patients who were not cured Figure 1. Sydney score of graded histological variables in the corpus (A) and antral (B) mucosa of 25 cured patients affected by atrophic body gastritis and H. pylori infection, before and 6 months after treatment. Results are shown as mean S.E.M. of atrophic body gastritis, neither gastrinemia nor Pepsinogenemia I were changed (data not shown). Serum IgG H. pylori values were signi cantly decreased on cured atrophic body gastritis patients (mean IU/L vs IU/L; P < ) whereas they tend to be increased, although not signi cantly, in not-cured patients (mean IU/L vs IU/L; P ˆ 0.079). Parietal cell antibodies were not modi ed by the treatment; in fact, only one of the initial 17 patients with the presence of circulating parietal cell antibodies showed a decrease of the titre, but with values still positive. Assessment at 1 year Acid secretion. The effect of H. pylori eradication on acid secretion was assessed on 15 atrophic body gastritis cured patients. Basal acid output was signi -

6 630 B. ANNIBALE et al. cantly increased 1 year after therapy (mean mmol/h vs mmol/h, P < 0.005; Figure 2A). In particular, nine patients showed a reappearance of basal acid output after the cure. Peak acid output values also were greatly affected by the eradication with a ve-fold increase in respect to the pre-treatment values (mean mmol/h vs mmol/h, P ˆ ; Figure 2B). With the exception of two patients, all subjects increased or restored their basal acid output capacity, and in six patients the peak acid output values were normalized (more than 13 mmol/h) (Figure 2B). Interestingly, two of the 15 subjects, having been cured for 1 year, developed heartburn and needed treatment with proton pump inhibitors. Their peak acid output values after the cure changed from 11.6 mmol/h to 53 mmol/h, and from 10 mmol/h to 41.9 mmol/h, respectively. A statistical inverse correlation was obtained (r ˆ , P < 0.05, Spearman rank correlation) between the histological score of corporal chronic in ltrate and peak acid output values in atrophic body gastritis cured patients, indicating that the higher the degree of chronic in ammatory in ltrate, the lower the ability of oxyntic mucosa to secrete acid. Gastric histology re-assessment. In order to verify the long-term effect of eradication, 15 cured patients repeated the gastric histological sampling 1 year after treatment. Figure 3 illustrates that both corporal and antral histological Sydney variables are overlap at 6 and 12 months. In particular, Figure 3A shows that the mean corporal scores for activity, chronic in ammation, atrophy and intestinal metaplasia overlap with the same variables at the 6-month follow-up and that only the rst two variables are signi cantly reduced by the eradication. The histological patterns of antral mucosa revealed the same behaviour (Figure 3B). Patterns of ECL cell hyperplasia. In order to evaluate the biological consequences of the reduced fasting hypergastrinemia induced by the eradication of H. pylori infection on 15 atrophic body gastritis patients we studied the patterns of ECL cell proliferation (the main gastric trophic target for the gastrin) before and 1 year after infection cure (Table 2). In eight patients (53%) a regression of severity of ECL hyperplastic change was observed, and in ve of them the ECL pattern returned to normal. No progression was observed and seven patients had a stable pattern. Fasting gastrin levels decreased signi cantly in the eight patients with regression of ECL pattern (mean pg/ml vs pg/ml, P ˆ ) while in the seven patients with no regression it decreased, although not signi cantly; (mean pg/ml vs pg/ml, P ˆ ). DISCUSSION Figure 2. Basal acid output (A) and peak-stimulated acid output (B) in 15 successfully cured patients with atrophic body gastritis, before and 12 months after eradication therapy. Bar indicates median. * P ˆ ; ** P ˆ The aim of this study investigating 35 consecutive patients with histologically proven atrophic body gastritis with H. pylori infection was to verify whether H. pylori eradication might modify gastric functional data (acid secretion and fasting gastrinemia) and

7 CURE OF H. PYLORI IN ATROPHIC BODY GASTRITIS 631 Figure 3. Sydney score of graded histological variables in the body (A) and antral (B) mucosa of 15 cured patients affected by atrophic body gastritis and H. pylori infection, before and 6 and 12 months after treatment. Results are shown as mean S.E.M. histological patterns (atrophy, intestinal metaplasia and ECL cell hyperplasia) of the patients. This study demonstrates that cure of H. pylori infection is effective in a signi cant proportion (78%) of atrophic body gastritis patients. H. pylori eradication results in a partial restoration of gastric acid secretion with consequent signi cant reduction in hypergastrinemia and the related effect on the fundic ECL cells, but without appreciable effect on body atrophy and intestinal metaplasia. The therapeutic regimen employed was the classic triple therapy based on bismuth administration: the use of antisecretory drugs was not indicated because of the lack of acid secretion in most patients. The good

8 632 B. ANNIBALE et al. Patient Gastrin before Tx Gastrin after Tx ECL pattern before tx ECL pattern after Tx Change Dysplasia Micronodular Regression Normal Normal Stable Normal Normal Stable Micronodular Linear Regression Linear Normal Regression Micronodular Micronodular Stable Micronodular Normal Regression Micronoluar Normal Regression Micronodular Micronodular Stable Micronodular Micronodular Stable Linear Normal Regression Micronodular Micronodular Stable Simple Normal Regression Micronodular Simple Regression Normal Normal Stable Table 2. Effect of H. pylori eradication on ECL cell proliferative pattern and fasting gastrin levels (pg/ml) in 15 atrophic body gastritis cured patients 1 year after the start of treatment eradication rate obtained in this consecutive population of atrophic body gastritis H. pylori-positive patients indicates that this regimen is a valid strategy for the cure of H. pylori infection as also already reported by us and others. 7, 9, 19 In fact, the rate of success was not different from that obtained in non-achlorhydric H. pylori-positive patients and, in spite of the high occurrence of minor side-effects, did not result in therapy withdrawal by any patient. 20 It is worth noting that in our patients the detection of infection was more dif cult than in those with duodenal ulcer or non-ulcer dyspepsia. In fact seven (20%) out of 35 atrophic body gastritis patients had no histological evidence of infection and only the use of subsidiary tests 7such as culture and Ab IgG allowed the diagnosis of H. pylori infection. This result con rms previous observations indicating that the `two positive tests' criterion for diagnosing H. pylori infection may be inappropriate for subjects with achlorhydria because bacterial colonization is reduced in the absence of gastric acid. 5, 21, 22 Among the 28 atrophic body gastritis patients with proven histological infection, however, the body H. pylori colonization was predominant and the grade 8 was conspicuous ( ) indicating another distinct feature of atrophic body gastritis patients in comparison with peptic ulcer disease patients in which the body H. pylori score colonization is less relevant. 14 This study documents that, although a partial recovery of acid secretion occurred in patients cured of atrophic body gastritis, body atrophy score and metaplastic change are not modi ed by the eradication, both parameters being unchanged in two consecutive histological evaluations (at 6 and 12 months after treatment). This nding was in contrast with the results of some studies, 10, 19, 22, 23 but was in agreement with 7, 24 those of other studies. The uncertainty of the criteria for de ning body atrophy may explain this discrepancy. For these criteria, in fact, the level of agreement between experienced pathologists is known to fall into poor statistical range. 16 A dense mononuclear cell in ltrate occupying the mucosa, in fact, may suggest sparse or completely lost glands, whereas only the coexistence or loss of the glands and replacement by intestinalized metaplastic epithelium represents true atrophy. 25 These strict criteria were not adopted in most previous studies, thus justifying the con icting results on the effect of H. pylori eradication on atrophy. 7, 19 Another possible source of bias is concerned with the extension of biopsy sampling in consideration of the patchy distribution of atrophic gastritis. In fact, most previous studies used only two biopsies for each type of mucosa (antral and corporal). 10, 19, 22, 23 In contrast, our bioptic procedure was based on a systematic sampling of the gastric mucosa and included four biopsies from the antrum and seven from the fundus/corpus, a methodology used to reduce the in uence of sampling error that was suggested by 25, 26 expert histopathologists. In this study we do not con rm the body atrophy reduction after the H. pylori cure found in our preliminary series of patients based on a small number of cases (n ˆ 7). 9 The present data were in agreement with those of circulating levels of Pepsinogen I, a reliable marker of chronic body atrophy, that were

9 CURE OF H. PYLORI IN ATROPHIC BODY GASTRITIS 633 greatly reduced before treatment but were not modi ed 27, 28 9after eradication. All these data do not support the suggestion of a glandular regenerative process of oxyntic mucosa. 19 In the antrum of the same atrophic body gastritis patients, H. pylori eradication induced also a clear-cut reduction in chronic in ammation that obviously represents the key effect of treatment, but no effect was observed again on atrophy or intestinal metaplasia, thus strengthening the corporal ndings. In our opinion the partial recovery of acid secretion observed after H. pylori cure in atrophic body gastritis patients is related to the reduction in chronic corporal in ammatory in ltrate. In fact, we have found a statistical inverse correlation between peak acid output values and grading of corporal chronic in ammatory in ltrate. These ndings are crucial since they could suggest a mechanism(s) by which H. pylori-associated in ammation of the stomach body inhibits acid secretion. Cytokinin induced by H. pylori infection (i.e. IL1b) were found to in uence acid secretion by reducing the ability of oxyntic mucosa to secrete acid in experimental conditions and its eradication, in non-body atrophic patients, is associated with a signi cant recovery of acid secretion. 29±33 All these data may account for how patients with H. pylori body atrophy can restore acid secretion when eradication of infection has occurred. In our cured atrophic body gastritis patients, the high signi cant reduction in chronic in ltrate and the complete disappearance of acute in ltrate in the body mucosa is in keeping with the work of Ruiz et al. who studied patients with multifocal atrophic gastritis. 34 They observed an improvement in acid secretion which coincided with an improvement in the in ammation and persistence of atrophy. The same relationship was also observed by Feldman et al. in dyspeptic patients. 28 Thus we suggested that the eradication of H. pylori is able to determine a functional recovery of those parietal cells still present but inhibited by the in ammatory H. pylori-related events. 35 The cure of infection determines a concomitant reduction in fasting hypergastrinemia as previously observed. 7, 9 The profound reduction in fasting gastrinemia induced an important biological consequence: 1 year after H. pylori cure the evolutionary patterns of ECL cell proliferation were slackened off, indicating that a partial restoration of the acid gastrin negative feed-back had occurred. This is the rst time that the trophic effect of hypergastrinemia on atrophic body gastritis patients has been reversed using a treatment not acting directly on gastrin cells. Until now, in fact, only surgical antrectomy or octreotide treatment had been demonstrated to counteract the trophic effect of hypergastrinemia on ECL cells of atrophic body gastritis patients. 36, 37 In conclusion, from this prospective investigation of patients with atrophic body gastritis, we documented that it is possible to cure H. pylori infection in most of the patients. Furthermore, by curing the infection it is possible to partially reverse the negative consequences of H. pylori infection on gastric corporal atrophy by increasing acid secretion and greatly reducing the hypergastrinemia. In these patients, therefore, H. pylori cure had another important advantage of removing a further potential risk for endocrine neoplastic change. 38 ACKNOWLEDGEMENTS This work was supported by grant no. 02/12/01/10, 1992±97 from the Italian Ministry for the University. We thank Ms Amelia Pasquali for her technical assistance. REFERENCES 1 Whittingham S, Mackay IR. Pernicious anemia and gastric atrophy. In: Rose NR, Mackay IR, eds. The Autoimmune Diseases. Orlando, Fl: Academic, 1985: 243±66. 2 Strickland RG, Mackay IRA. Reappraisal, signi cance of chronic atrophic gastritis. Dig Dis 1973; 18: 426±40. 3 Appelman HD. Gastritis: terminology etiology, clinicopathological correlations: Another biased view. Human Pathol 1994; 25: 1006±19. 4 Maton PN. Gastrinoma and Other Hypergastrinemic Syndromes. In: Walsh JH, Dockrey GJ, eds. Gut Peptides: Biochemistry and Physiology. New York: Raven Press, 1994: 675± Fong TL, Dooley CP, Dehesa M, et al. Helicobacter pylori infection in pernicious anemia: a prospective controlled study. Gastroenterology 1991; 100: 328±32. 6 Karnes WE, Samloff IM, Siurala M, et al. Positive serum antibody, negative issue staining for Helicobacter pylori in subjects with atrophic body gastritis. Gastroenterology 1991; 101: 167±74. 7 El-Omar EM, Oien K, El-Nujumi A, et al. Helicobacter pylori infection, chronic gastric acid hyposecretion. Gastroenterology 1997; 113: 15±24. 8 Kuipers EJ, Uyterlinde AM, Pena AS, et al. Long-term sequelae of Helicobacter pylori gastritis. Lancet 1995; 345: 1525±28. 9 Annibale B, Marignani M, Azzoni C, et al. Atrophic body gastritis: distinct features associated with Helicobacter pylori infection. Helicobacter 1997; 2: 57± Tucci A, Biasco G, Paparo GF. Effect of eradication of Helicobacter pylori in patients with fundic atrophic gastritis. N Eng J Med 1997; 336: 957±58.

10 634 B. ANNIBALE et al. 11 Bordi C, Annibale B, Azzoni C, et al. Endocrine cell growths in atrophic body gastritis. Critical evaluation of a histological 10 classi cation. J Pathol 1997; 182: 339± Weinstein WM. Gastritis and gastropathies. In: Sleisenger MH, 11 Fordtran JS, eds. Gastrointestinal Disease, 5th edn. Philadelphia: Saunders, 1993: 545± Correa P. Chronic gastritis: a clinic-pathological classi cation. Am J Gastroenterol 1988; 83: Dixon MF, Genta RM, Yardley JH, et al. Classi cation and grading of gastritis. The updated Sydney system. Am J Surg 12 Pathol 1996; 20: 1161± Solcia E, Bordi C, Creutzfeldt W, et al. Histopathological classi cation of nonantral gastric endocrine growths in man. Digestion 1988; 41: 185± Genta RM, Ginger ML, Graham DY. Changes in the gastric mucosa following eradication of Helicobacter pylori. Mod Pathol 1993; 6: 281±9. 17 Delle Fave G, Kohn A, de Magistris L, et al. Effects of bombesin on gastrin and gastric acid secretion in patients with duodenal ulcer. Gut 1983; 24: 231±5. 18 Annibale B, de Magistris L, Corleto V, et al. Zollinger±Ellison syndrome and antral G-cell hyperfunction in patients with resistant duodenal ulcer disease. Aliment Pharmacol Ther 1994; 8: 87± Tucci A, Poli L, Tosetti G, et al. Reversal of fundic atrophy after eradication of Helicobacter pylori. Am J Gastroenterol 1998; 93: 1425± Unge P. What other regimens are under investigation to treat Helicobacter pylori infection? Gastroenterology 1997; 113: S131± Faisal MA, Russell RM, Samloff IM, et al. Helicobacter pylori infection and atrophic gastritis in the elderly. Gastroenterology 1990; 99: 1543±4. 22 Oberhuber G, Wuendisch T, Rappel S, et al. Signi cant improvement of atrophy after eradication therapy in atrophic body gastritis. Pathol Res Pract 1998; 194: 609± Haruma K, Mihara M, Okamoto E, et al. Eradication of Helicobacter pylori increases gastric acidity in patients with atrophic gastritis of the corpusðevaluation of 24-h ph monitoring Aliment Pharmacol Ther 1999; 13: 155± Satoh K, Kimura K, Takimoto T, et al. A follow-up study of atrophic gastritis and intestinal metaplasia after the eradication of Helicobacter pylori. Helicobacter 1998; 3: 236± Genta RM. Helicobacter pylori, in ammation, mucosal damage and apoptosis: Pathogenesis and de nition of gastric atrophy. Gastroenterology 1997; 113: S51±5. 26 Domellof L. Reversal of gastric atrophy after Helicobacter pylori eradication: is it possible or not. Am J Gastroenterol 1998; 93: 1407±8. 27 Samloff IM, Varis K, Ihamaki T, et al. Relationship among serum pepsinogen I, serum pepsinogen II and gastric mucosal history. A study in relatives of patients with pernicious anemia. Gastroenterology 1982; 83: 204±9. 28 Feldman M, Cryer B, McArthur CE, et al. Effects of ageing and gastritis on gastric acid and pepsin secretion in human: a 14 prospective study. Gastroenterology 1996; 110: 1043± Kondo S, Shimura Y, Kanayama S, et al. Interleukin 1 beta inhibits gastric histamine secretion and synthesis in rat. Am J Physiol 1994; 267: G966± Yasunaga Y, Shinomura Y, Kanayama S. et al. Mucosal interleukin 1 beta production and acid secretion in enlarged fold gastritis. Aliment Pharmacol Ther 1997; 11: 801±9. 31 Taguchi Y, Kaito M, Gabazza EC, et al. Helicobacter pylori inhibits the secretory activity of gastric parietal cells in patients with chronic gastritis. Scand J Gastroenterol 1997; 32: 656± Furuta T, Baba S, Takashima M, et al. Effect of Helicobacter pylori infection on gastric juice ph. Scand J Gastroenterology 1998; 33: 357± Feldman M, Cryer B, Lee E. Effects of Helicobacter pylori gastritis on gastric secretion in healthy human beings. Am J Physiol (Gastroenterol) 1998; 274: G1011±7. 34 Ruiz B, Correa P, Fonthman ET, et al. Antral atrophy, Helicobacter pylori colonization, and gastric ph. Am J Clin Pathol 1996; 105: 96± McGowan CC, Cover TL, Blaser MJ. Helicobacter pylori and gastric acid: biological and therapeutics implications. Gastroenterology 1996; 110: 926± D'Adda T, Annibale B, Delle Fave G, et al. Oxyntic endocrine cells of hypergastrimenic patients. Differential responses to antrectomy or octreotide. Gut 1996; 38: 668± Ferraro G, Annibale B, Marignani M, et al. Effectiveness of octreotide in controlling fasting hypergastrinemia and related enterochromaf n-like cell growth. J Clin Endocr Metab 1996; 81: 677± Sjoblom SM, Sipponen P, Jarvinen H. Gastroscopic follow-up of pernicious anemia patients. Gut 1993; 34: 28±32.