0002-9270/93/881I-187003.00/0 THE AMERICAN JOURNAL OF GASTROENTEROLOGY Copyright © 1993 by Am. Coll. of Gastroenterology Il Vol. 88, No. 11, 1993 Printed in U.S.A. Helicobacter pylori Causes Hyperproliferation of the Gastric Epithelium: Pre- and Post-Eradication Indices of Proliferating Cell Nuclear Antigen Fernando Brenes, M.D., Bernardo Ruiz, M.D., Pelayo Correa, M.D., Fred Hunter, M.D., Thiruvengadam Rhamakrishnan, M.D., Elizabeth Fontham, Ph.D., and Tei-Ying Shi, M.D. Department of Pathology, and Department of eAcine, Louisiana State University Medical Center, New Orleans, Louisiana Objective: To assess the effect of Helicobacter pylori (HP) eradication on the proliferation of the gastric epithelium by the expression of the proliferating cell nuclear antigen (PCNA). Methods: Alcohol-fixed gas- tric biopsies taken before and after treatment for HP were immunostained with the PC-10 anti-PCNA mono- clonal antibody and the labeling index was determined with an image analysis system. Results: The mean PCNA-labeling index (LI) of 16 patients who remained HP positive did not change significantly (18.95 ± 1.71 on first visit vs. 17.96 ± 1.91 on second visit, mean ± SEM). The mean PCNA LI of 31 patients who cleared HP was reduced significantly (19.95 ± 1.77 on first visit vs. 14.13 ± 1.29 on second visit, p < 0.001). Patients who were positive for HP at both first and second visit showed a significantly higher PCNA LI than normal control biopsies (13.05 ± 1.70) (p < 0.05). Conclusions: These results indicate that the gastric mucosa infected with HP is in a state of hyperprolifer- ation. Patients who cleared HP showed a significant histopathological improvement, reflected in the reduc- tion in number of polymorphonuclear and mononuclear cell infiltrates, and a reduction in the amount of super- ficial mucosa damage. Prolonged hyperproliferation of the gastric epithelium exerted by HP infection could be a major factor for human gastric carcinogenesis. INTRODUCTION Gastritis associated with Helicobacter pylori (HP) infection is among the most frequent chronic infectious diseases in the world. Its estimated prevalence in adults varies from approximately 30% to near 100% in differ- ent populations (1-3). According to some authors, the syndrome most frequently related to HP infection is non-ulcer dyspepsia (4, 5). The bacterium is also con- sidered a contributing cause of peptic ulcer and gastric carcinoma (2, 6-11). Because the bacterium colonizes the gastric lumen Received Apr. 19, 1993; accepted July 1, 1993. but does not invade the tissues, the mechanisms of tissue damage are the subject of scientific scrutiny. Cellular damage due to bacterial cytotoxin (12) and to ammonia produced by its potent urease have been implicated (13). The mechanism of its putative carcin- ogenic effect is unknown. This report explores the hypothesis that excessive cell replication, a factor of major relevance in carcinogenesis (14, 15), may be caused in the gastric mucosa by HP infection. Cell replication is assessed in this study by the effect of HP eradication on the expression of proliferating cell nu- clear antigen (PCNA). MATERIALS AND METHODS Fifty-eight patients referred for upper gastrointestinal endoscopy at Charity Hospital in New Orleans, with a clinical and endoscopic diagnosis of non-ulcer dyspep- sia, were included in this study. Each agreed to partic- ipate in a longitudinal study of the effects of HP on gastric histopathology, approved by the Louisiana State University Institutional Review Board. Eleven patients who were found to have normal gastric mucosa were negative for HP. These subjects had only one visit and were used as controls. The rest of the patients had gastroscopic biopsies positive for HP on rapid urease test (16) and were invited to participate in the study. The subjects were assigned treatments after the first visit as follows: 13 received no specific treatment for HP, 23 received triple therapy consisting of: 1) bismuth subsalicylate (262 mg qid for 28 days), 2) metronida- zole (500 mg qid on days 18-28), and 3) amoxycyllin (500 mg qid on days 8-24); and 11 patients received triple therapy with nitrofurantoin (100 mg qid on days 8-17) instead of amoxycyllin. Endoscopies and biopsies were repeated on the second visit as follows. In those who received triple therapy, the mean time from ter- mination of treatment to second biopsy was 3 days (range 12 h to 23 days), with a median of 1 day. In this group, 10 patients had a second biopsy 12 h after finishing treatment, and 24 patients had a second bi- 1870 AJG — November 1993 HP-CAUSED HYPERPROLIFERATION OF GASTRIC EPITHELIUM 1871 opsy 36 h or more after finishing treatment. Patients who received no specific treatment for HP had a second biopsy after a mean of 147 days (range 56-226 days), and a median of 168 days. Five antral biopsies and one corporal biopsy were taken on each visit. One antral biopsy, from the lesser curvature, was used for the rapid urease test for HP. Two antral biopsies (greater and lesser curvature) were fixed in 95% alcohol, and two (greater and lesser cur- vature) were fixed in 10% buffered formalin. The cor- poral biopsy was also fixed in 10% buffered formalin. Tissues were dehydrated and paraffin embedded, sec- tioned at 3 Am and placed on poly-L-lysine (Sigma Chemical Co.) -coated glass slides. The formalin-fixed tissues were stained with hematoxylin and eosin, periodic acid-Schiff-Alcian blue (PAS-AB) pH 2.5 (17), and the Steiner modification of Warthin-Starry staining for HP (18). The 95% alcohol fixed tissues were immunostained for PCNA with PC10 monoclonal antibody (Dako Cor- poration), by a microwave-assisted streptavidin-biotin peroxidase method, and the Fisher Capillary System (Fisher Scientific). This method allowed us to immu- nostain 60 slides at the same time, under the same conditions for antibody incubation. Briefly, sections were deparaffinized and rehydrated. Endogenous per- oxidase was blocked with 3% hydrogen peroxide for 10 s in a microwave oven at high power (1440 watts). After washing in deionized water, sections were boiled 6 min in deionized water in the microwave oven at high power and allowed to stand for 3 min (19). Sections were placed in the Capillary System slide holder and washed in phosphate-buffered saline (PBS) (pH 7.2) with 0.01% Tween 20 (Sigma) 3 times for 10 s, then were drained over a thick paper pad. All the antibody incubations were done in the microwave oven at 30% power and allowed to stand for 3 min. PC10 monoclonal antibody was incubated for 14 min. Secondary biotinylated goat antimouse immunoglobulins (Dako) was incubated for 7 min. Streptavidin-peroxidase complex (Dako) was incubated for 7 min. Diaminobenzidine (Sigma) was used at 0.05% in PBS for 10 min at room temperature. Sections were lightly counterstained with Harris he- matoxylin, dehydrated, and mounted. Negative con- trols included substitution of primary antibody with PBS or normal mouse serum. A Joyce-Loebl Magiscan (London, England) image analysis system was used to assess the PCNA nuclear labeling. PCNA was evaluated in areas corresponding to gastric antrum glandular necks at x400, using a static 560-nm filter. Seven glandular necks were se- lected at random in each specimen to ensure a mini- mum of 250 total nuclei. System thresholding was adjusted so only those nuclei stained a strong dark brown were counted as positive. Then, the threshold was adjusted to count the total nuclei in the selected areas. The mean number of nuclei counted per speci- men was 508 (range 267-752, median 526). The PCNA labeling index (PCNA LI) was defined as a percentage of the PCNA-positive nuclei over the total nuclei in glandular neck regions. Biopsies were diagnosed histopathologically accord- ing to previously defined criteria (20) in diffuse antral gastritis (DAG) and multifocal atrophic gastritis (MAG) on formalin-fixed and hematoxylin and eosin-stained sections of corpus and antrum. DAG is characterized by a dense infiltrate of lymphocytes and plasma cells occupying the full thickness of the antral mucosa. The infiltrate expands the lamina propria, separating the gastric glands, and lymphoid follicles may be promi- nent; this entity is most often related to the duodenal ulcer syndrome. MAG corresponds to multifocal glan- dular atrophy accompanied by an inflammatory mono- nuclear infiltrate, and is located in antrum and body. Intestinal metaplasia is often found associated with MAG. A semiquantitative assessment of the histopath- ologic changes was made by two observers; discrepan- cies were resolved by consensus after joint review. The presence of polymorphonuclear and mononuclear cells in the lamina propria and glands was evaluated as 0 = normal or very rare, 1 = a few or equivocal, 2 = moderate number, and 3 = marked increase. Damage to superficial and foveolar epithelia, along with mucus depletion, was evaluated in both hematoxylin and eosin and PAS-AB, and given the following scores: 0 = nor- mal superficial mucosa, 1 = mild superficial mucosa damage, 2 = moderate superficial mucosa damage, and 3 = marked superficial mucosa damage. The presence of HP was scored accordingly, 0 = negative, 1 = small number, 2 = moderate number, and 3 = abundant. A global histological score, adding all the previous param- eter scores except for HP, was used for comparison. All of the slides were evaluated blindly, without previous knowledge of the histological diagnosis, treat- ment, or treatment outcome for HP. PCNA LI was evaluated only in non-metaplastic areas. Student's t test (two-tailed) was used to evaluate pre- and post-treatment differences in PCNA LI. Compari- son of histopathological scores was performed by Wil- coxon's rank test (two-tailed). RESULTS The study included 39 females and 19 males, with a mean age of 47.7 yr (range 20-70). Histological diag- nosis included 40 patients with MAG, seven with DAG, and 11 normal gastric biopsies used as controls. Immunostaining of PCNA, with a diffuse granular pattern, was localized in the nuclei. The staining inten- sity was variable. In the normal cases, strong labeling was observed only in the glandular necks (Fig. 1A). 1 '•:-7:14 :, .• .-. ;:.• ' 7 T,,:: - ! .,., r, ' fi • pi' • i':' ' " ' • , m 4 • : • ' c . IX • ..,.(4 * 411., ` v'''' ' ,. ,... • IF a' • 11' 'a."-.^. v** ,;• " a • , . . ." ' 4' ; • .. . • 1 t ( ft .R. • .1 +10 f;* : . ''' ' • . IN , . ' 9 1 "•:. ' • * ib • •' ' 4. a I . .4% , • . ' 4,.. • ' ' i. , e 4 ... 1.5 ' t #' * i. f r • ta •Fr a - v • ,,,,. A ,.. L. ...A t 4, 4 a s 1872 BRENES et al. FIG. 1. PCNA immunostaining of the gastric mucosa. A, normal gastric mucosa expressing PCNA in the nuclei of cells of the glandular necks (x200). B, MAG, showing an increased number of PCNA- positive nuclei and extension to superficial and deep areas of the mucosa (x200). AJG - Vol. 88, No. 11, 1993 Some of these cases had weak staining in a few cells of the superficial mucosa. In chronic gastritis, strong nu- clear PCNA immunostaining was observed mainly in glandular necks (Fig. 1B), but in some cases, few stained nuclei were seen in the deep or superficial areas of the mucosa. The PCNA stain was evaluated only in glan- dular necks. The few positively stained nuclei in other layers were insufficient for adequate evaluation. The 11 patients with normal gastric histology and negative for HP showed a mean PCNA LI of 13.05 ± 1.70 (Table 1). On the second visit, of the 47 patients initially positive for HP, 31 cleared and 16 persisted with the infection. In patients who remained HP posi- tive, the PCNA LI did not change significantly (Table 1). The mean PCNA LI for those who cleared HP decreased significantly (p < 0.001). Patients who re- mained positive for HP, on both the first and second visit, showed a significantly higher PCNA LI than nor- mal control biopsies (p < 0.05). However, patients who cleared HP infection showed a significantly higher PCNA LI only on the first visit, in relation to normal gastric mucosa (p < 0.01). According to histopathological diagnosis, patients with MAG, whose HP cleared, showed a significant reduction in PCNA LI (n = 19, 22.10 ± 2.61 vs. 15.09 ± 1.33, p < 0.05), and those that did not clear showed no difference (n = 16, 20.71 ± 2.65 vs. 19.51 ± 1.97). In DAG, five patients cleared HP and showed a tend- ency to maintain similar PCNA LI (17.85 ± 3.52 vs. 16.90 ± 5.69); two patients did not clear, but were not sufficient for statistical analysis. Histopathological parameters were evaluated by the system of scores described above. In patients who cleared HP, there was a significant reduction of the inflammatory, infiltrate, reflected in the scores for pol- ymorphonuclear and mononuclear cell infiltration (p < 0.001) (Table 2). A significant improvement in the mucus production (p < 0.001) was observed in those patients who cleared HP. These improvements after HP clearance were reflected in the significant reduction of the global histopathological score (p < 0.001). Histo- pathological parameters in patients who remained pos- itive did not change significantly. Table 3 displays the PCNA LI on the second visit, according to the histologic parameters studied and HP TABLE 1 PCNA Labeling Index* According to Study Results First Visit Second Visit p HP cleared (n = 31) 19.95 ± 1.77 14.13 ± 1.29 <0.001 HP persisted (n = 16) 18.95 ± 1.71 17.96 ± 1.91 NSt Nl. histology, HP -ye (n = 11) 13.05 ± 1.70 * Mean ± SEM. t NS, not statistically significant. AJG - November 1993 HP-CAUSED HYPERPROLIFERATION OF GASTRIC EPITHELIUM 1873 TABLE 2 Comparison of Scores for Histological Parameters and Helicobacter pylori Status Parameter HP Cleared (n = 31) HP Persisted (n = 16) First visit Second visit First visit Second visit Median score p Median score p Sup. mucosa dam. 2 0 <0.001 1 2 NS PMN cells 2 0 <0.001 2 2 NS MN cells 2 1 <0.001 2 2 NS Histol. score 7 3 <0.001 6 7 NS Abbreviations: Sup. mucosa dam, superficial mucosa damage; PMN, polymorphonuclear; MN, mononuclear; NS, not statistically significant. TABLE 3 PCNA LI* According to Histologic Parameters and HP Status on Second Visit All Observations (n = 58) HP Positive (n = 16) HP Negative (n = 31) PMN infiltrate Absent or mild (46) 13.91 ± 0.99 (4) 14.53 ± 3.40 (31) 14.13 ± 1.29 Moderate or severe (12) 19.11 ± 2.26 (12) 19.11 ± 2.26 (0) p = 0.024 p = 0.316 ISS MN infiltrate Absent or mild (33) 13.90 ± 1.25 (4) 14.53 ± 3.40 (18) 14.27 ± 1.97 Moderate or severe (25) 16.41 ± 1.41 (12) 19.11 ± 2.26 (13) 13.93 ± 1.51 p = 0.190 p = 0.316 p = 0.898 Sup. mucos. dam. No depletion (45) 13.90 ± 0.99 (5) 12.04 ± 2.41 (29) 14.54 ± 1.34 Some depletion (13) 18.73 ± 2.23 (11) 20.65 ± 2.15 (2) 8.18 ± 0.78 p = 0.031 p = 0.031 ISS Results are expressed as mean ± SEM. Number of patients in parentheses. PMN, polymorphonuclear cells; MN, mononuclear cells; Sup. mucos. dam., superficial mucosa damage; ISS, insufficient sample size. status. If the whole study group is considered, those patients with a moderate or severe polymorphonuclear cell infiltrate and those with superficial mucosa damage showed a significantly higher PCNA LI. In the whole group, and in those positive for HP, the greater intensity of the mononuclear infiltrate was associated with a tendency to a higher LI, but it did not reach statistical significance. After clearance of HP, the PCNA LI did not differ significantly according to the intensity of mononuclear infiltrate. In the group that cleared HP, both the polymorphonuclear infiltrate and the superfi- cial mucosa damage disappeared, making impossible their independent evaluation. DISCUSSION The value of PCNA LI in assessing cell proliferation is supported by studies of carcinoma of the stomach showing a good correlation with S + G2M phase of flow cytometric analysis (21) and with Ki-67 (22). In other systems, PCNA LI has shown a good correlation with other indexes of cell proliferation such as [3H]thymidine (23, 24), and bromodeoxyuridine (BrdUrd) (24, 25). PCNA is mainly expressed in late G1 phase, during the S phase and early G2 (26), and it can be detected at low levels in cells during the cell cycle. PCNA LI, compared with BrdUrd and flow cytometry, has shown a higher percentage of proliferating cells (22, 23, 27). One explanation offered for these results is the long half-life of PCNA (about 20 h) (28). Image analysis thresholding has allowed us to select only those nuclei that are strongly positive, which has been reported to have a better correlation with BrdUrd and flow cytom- etry cell proliferation indices (29, 30). PCNA reportedly had a better correlation with [3H]thymidine labeling when alcohol-fixed tissues were used (23, 27, 29, 30). Cell proliferation studies in chronic gastritis have shown a high number of proliferating cells (31, 32), consistently higher than the normal gastric mucosa. There have been similar findings on colonic mucosa, in which a state of hyperproliferation is present in precancerous conditions such as familial polyposis, ad- enomas, and ulcerative colitis (33-36). Our study shows a significant reduction of PCNA- labeled cells, indicative of cell replication, after clear- ance of HP. The post-clearance indices are very close to those of the normal gastric mucosa. These results indicate that the infected gastric mucosa is in a state of hyperproliferation. Histopathological parameters were improved after 1874 BRENES et al. HP eradication in our study. Both acute and chronic inflammatory infiltrates were reduced in significant numbers. Superficial mucosa damage and mucus de- pletion were significantly attenuated after HP clearance. These changes corroborate those observed by others (37-40) and confirm the damage inflicted by HP on the gastric mucosa. Changes in the chronic (lympho- cytic) infiltrate need to be studied for a longer term. Given the chronicity of the gastritis, it is probable that the state of hyperproliferation is present for many years in most patients. Because HP infection increases gastric cancer risk, this situation fits the hypothesis proposed by Ames and others (41, 42) that in a hyper- proliferative state, endogenous ("spontaneous") muta- tions may induce genetic lesions which accumulate and eventually might lead to neoplastic transformation (43, 44). Helicobacter-associated gastritis provides a model to use to examine this hypothesis, because persistence of hyperproliferation over a long period of time is associated with an increased risk of carcinoma (2, 8, 6, 45, 46). According to the Ames hypothesis, endogenous mutations per se, without external mutagens or carcin- ogens, may be sufficient to induce neoplastic changes in tissues in a state of hyperproliferation (42). In populations at high risk of gastric carcinoma, the prevalence of Helicobacter-associated gastritis is very high (2, 8, 6, 45, 46). The type of gastritis predominant in such populations is the so called multifocal atrophic gastritis (MAG), as it happens in the New Orleans blacks, who predominate in our study (12). Mucosal atrophy is followed by intestinal metaplasia and, even- tually, dysplasia and carcinoma (47). Helicobacter in- fection is very prevalent and severe in such populations, which strongly suggests a causative role for the bacteria in carcinogenesis (9). HP, however, is also prevalent in DAG, which is the predominant type of gastritis in populations at low gastric cancer risk, such as the white middle-class Amer- icans (9). DAG is also the predominant type of gastritis seen in children. Recent studies in Costa Rica compared the prevalence of gastritis and Helicobacter infection in children from the central plateau (who have some of the highest gastric cancer risks on record), with those from the coastal lowlands, where the gastric cancer risk is low (48). In both populations the prevalence of HP- associated gastritis is very high: around 70% in children of 7-12 yr of age. Because the gastritis in the coastal lowlands is not associated with higher cancer risk, it follows that HP is not a sufficient cause of gastric cancer. Other factors, presumably present in the plateau highlands, are needed and may probably be more effec- tive in the presence of the hyperproliferative state brought about by HP infection. The same pattern of high prevalence of HP-associated chronic gastritis but low gastric cancer risk is observed in Africa (3). These AJG - Vol. 88, No. 11, 1993 observations suggest that endogenous mutations per se are not sufficient to increase the cancer rates to epi- demic proportions. Our studies indicate that prolonged hyperprolifera- tion of the gastric epithelium is a major mechanism by which HP exerts a contributing causal influence in human gastric carcinogenesis. ACKNOWLEDGMENTS This work was supported by grants R01-AI-25590, PO 1 -CA-28842 from the National Institute of Health and the National Cancer Institute. F. Brenes is a Fellow of the National Cancer Institute, and is supported by the Costa Rican Health System (CCSS), Costa Rican National Research and Technology Council (CONI- CIT), and University of Costa Rica. Reprint requests and correspondence: Pelayo Correa, MD, Loui- siana State Medical Center, Department of Pathology, 1901 Perdido Street, New Orleans, LA 70112. REFERENCES 1. Correa P, Fox J, Fontham E, et al. Helicobacter pylori and gastric carcinoma. Serum antibody prevalence in populations with con- trasting cancer risks. Cancer 1990;66:2569-74. 2. Parsonnet J, Friedman GD, Vandersteen DP, et al. Helicobacter pylori infection and the risk of gastric carcinoma. N Engl J Med 1991;325:1170-1. 3. Megraud F, Brassens-Rabbe MP, Denis F, et al. Seroepidemiol- ogy of Campylobacter pylori infection in various populations. J Clin Microbiol 1989;27:1870-3. 4. Graham DY. Pathogenic mechanisms leading to Helicobacter pylori-induced inflammation. Eur J Gastroenterol Hepatol 1992;4(suppl 2):S9-S16. 5. Cover TL, Blaser MJ. Helicobacter pylori and gastroduodenal disease. Annu Rev Med 1992;43:135-45. 6. Nomura A, Stemmermann GN, Chyou PH, et al. Helicobacter pylori infection and gastric carcinoma among Japanese Ameri- cans in Hawaii. N Engl J Med 1991;325:1132-6. 7. Blaser MJ. Gastric Campylobacter-like organisms, gastritis and peptic ulcer disease. Gastroenterology 1987;93:371-83. 8. Forman D, Newell DG, Fullerton F, et al. Association between infection with Helicobacter pylori and risk of gastric cancer: Evidence from a prospective investigation. Br Med J 1991;302:1302-5. 9. Correa P, Ruiz B. Campylobacter pylori and gastric cancer. In: Rathbone BJ, Heatley RV, eds. Campylobacter pylori and gastro- duodenal disease. Oxford: Blackwell Scientific Publications, 1989:139-45. 10. Correa P. Is gastric carcinoma an infectious disease? N Engl J Med 1991;325:1170-1. 11. The Eurogast Study Group. An international association between Helicobacter pylori infection and gastric cancer. Lancet 1993;341:1359-62. 12. Fox JG, Correa P, Taylor NS, et al. High prevalence and persist- ence of cytotoxin positive Helicobacter pylori strains in a popu- lation with high prevalence of atrophic gastritis. Am J Gastroen- terol 1992;87:1554-60. 13. Megraud F, Neman-Simha V, Briigmann D. Further evidence of the toxic effect of ammonia produced by Helicobacter pylori urease on human epithelial cells. Infect Immun 1992;60:1858- 63. 14. Lipkin M. Biomarkers of increased susceptibility to gastrointes- tinal cancer: New application to studies of cancer prevention in human subjects. Cancer Res 1988;48:235-45. 15. Biasco G, Paganelli GM, Miglioli M, et al. Cell proliferation AJG - November 1993 HP-CAUSED HYPERPROLIFERATION OF GASTRIC EPITHELIUM 1875 biomarkers in the gastrointestinal tract. J Cell Biochem 1992;(suppl 16G):73-8. 16. Ruiz B, Janney A, Diavolitsis S, et al. One-minute test for Campylobacter pylori. Am J Gastroenterol 1989;84:202. 17. Luna LG. PAS-Alcian blue method for mucosubstances, pH 2.5. In: Luna LG, ed. Manual of histologic staining methods of the Armed Forces Institute of Pathology, 3rd ed. New York: McGraw Hill Book Co., 1968:168-9. 18. Garvey W, Fathi A, Bigelow F. Modified Steiner for the dem- onstration of spirochetes. J Histotechnol 1985;8:15-7. 19. Shi SR, Key ME, Kalra KL. Antigen retrieval in formalin-fixed, paraffin-embedded tissues: An enhancement method for immu- nohistochemical staining based on microwave oven heating of tissue sections. J Histochem Cytochem 1991;39:741-8. 20. Correa P. Chronic gastritis: A clinico-pathological classification. Am J Gastroenterol 1988;83:504-9. 21. Jain S, Filipe MI, Hall PA, et al. Prognostic value of proliferating cell nuclear antigen in gastric carcinoma. J Clin Pathol 1991;44:655-9. 22. Rosa JC, Mendes R, Filippe MI, et al. Measurement of cell proliferation in gastric carcinoma: Comparative analysis of Ki- 67 and proliferative cell nuclear antigen (PCNA). Histochem J 1992;24:93-101. 23. Richter F, Richter A, Yan K, et al. Cell proliferation in rat colon measured with bromodeoxyuridine, proliferating nuclear antigen and [3H]thymidine. Cancer Epidem Biomed Prey 1992;1:561-6. 24. Yamada K, Yoshitake K, Sato M, et al. Proliferating cell nuclear antigen expression in normal, preneoplastic, and neoplastic co- lonic epithelium of the rat. Gastroenterology 1992;103:160-7. 25. Sarraf ES, McCormick CSF, Brown GR, et al. Proliferating cell nuclear antigen immunolocalization in gastro-intestinal epithelia. Digestion 1991;50:85-91. 26. Celis JE, Celis A. Cell cycle dependent variations in the distri- bution of the nuclear protein cyclin proliferating cell nuclear antigen in cultured cells: Subdivision of S phase. Proc Natl Acad Sci USA 1985;82:3262-6. 27. Garcia LG, Coltrera MD, Gown AM. Analysis of proliferative grade using anti-PCNA/cyclin monoclonal antibodies in fixed, embedded tissues. Comparison with flow cytometric analysis. Am J Pathol 1989;134:733-9. 28. Bravo R, Macdonald-Bravo H. Existence of two populations of cyclin/proliferating cell nuclear antigen during the cell cycle: Association with DNA replication sites. J Cell Biol 1987;105:1549-54. 29. Monaghan P, Perusinghe NP, Nicholson RI, et al. Growth factor stimulation of proliferating cell nuclear antigen (PCNA) in hu- man breast epithelium in organ culture. Cell Biol Int Rep 1991;15:561-70. 30. Benjamin DR. Proliferating cell nuclear antigen (PCNA) and pediatric tumors: Assessment of proliferative activity. Pediatr Pathol 1991;11:507-19. 31. Lipkin M, Correa P, Mikol Y, et al. Proliferative and antigenic modifications of human epithelial cells in chronic atrophic gas- tritis. J Nall Cancer Inst 1985;75:613-9. 32. Biasco G, Paganelli GM, Brillanti S, et al. Cell renewal and cancer risk of the stomach: Analysis of cell proliferation kinetics in atrophic gastritis. Acta Gastroenterol Belg 1989;52:361-6. 33. Bleiberg J, Mainguet P, Galand P. Cell renewal in familial polyposis: Comparison between polyps and adjacent healthy mucosa. Gastroenterology 1972;62:240-.5. 34. Biasco G, Lipkin M, Minarini A, et al. Proliferative and antigenic properties of the rectal cells in patients with ulcerative chronic colitis. Cancer Res 1984;44:5450-4. 35. Terpstra OT, Blankestein M van, Dees J, et al. Abnormal pattern of cell proliferation in the entire colonic mucosa of patients with colon adenoma or cancer. Gastroenterology 1987;92:704-8. 36. Riddell RH, Goldman H, Ransohoff DF, et al. Dysplasia in inflammatory bowel disease: Standardized classification with pro- visional clinical applications. Hum Pathol 1983;14:931-51 37. Rauws EAJ, Langenberg W, Houthoff HJ, et al. Campylobacter pyloridis-associated chronic active antral gastritis. A prospective study of its prevalence and the effects of antibacterial and antiul- cer treatment. Gastroenterology 1988;94:33-40. 38. Kawano S, Tsujii M, Nagano K, et al. Different effect of Heli- cobacter pylori on the human gastric antral and body mucosa intracellular mucin. Scand J Gastroenterol 1990;25:997-W03. 39. Valle J, Seppala K, Sipponen P, et al. Disappearance of gastritis after eradication of Helicobacter pylori. Scand J Gastroenterol 1991;26:1057-65. 40. Genta RM, Lew GM, Graham DY. Changes in the gastric mucosa following eradication of Helicobacter pylori. Modern Pathol 1993;6:281-9. 41. Ames BN, Gold LS. Too many rodent carcinogens: Mitogenesis increases mutagenesis. Science 1990;249:970-1. 42. Preston-Martin S, Pike Malcom C, Koss RK, et al. Increased cell division as a cause of human cancer. Cancer Res 1990;50:7415- 21. 43. Cohen SM, Ellwein LB. Cell proliferation in carcinogenesis. Science 1990;249:1007-11. 44. Cohen SM, Purtillo DT, Ellwein LB. Pivotal role of increased cell proliferation in human carcinogenesis. Mod Pathol 1991;4:371-82. 45. Talley NJ, Zinsmeister AR, Weaver A, et al. Gastric adenocar- cinoma and Helicobacter pylori infection. J Natl Cancer Inst 1991;83:1734-9. 46. Sipponen P, Kosunen TU, Valle J, et al. Helicobacter pylori infection and chronic gastritis in gastric cancer. J Clin Pathol 1992;45:319-23. 47. Correa P. A human model of gastric carcinogenesis. Cancer Res 1988;48:3554-60. 48. Sierra R, Munoz N, Pella AS, et al. Antibodies to Helicobacter pylori and pepsinogen levels in children from Costa Rica: Com- parison of two areas with different risks for stomach cancer. Cancer Epid Biomed Prey 1992;1:449-59.