Mucosal change of the stomach with low-grade mucosa-associated lymphoid tissue lymphoma after eradication of Helicobacter pylori:Follow-up study of 48cases
Shahnaz Beguma, Toshiaki Sanoa, Hideko Endoa, Hitoshi Kawamatab and Yoshihito Urakamic

aDepartment of Pathology, The University of Tokushima School of Medicine, Tokushima, Japan; bDepartment of Oral and Maxillofacial Surgery, The University of Tokushima School of Dentistry, Tokushima, Japan;and cUrakami Gastro Clinic, Tokushima, Japan

Abstract: Low-grade mucosa-associated lymphoid tissue (MALT) lymphoma of the stomach has been demonstrated to be closely linked to Helicobacter pylori (H. pylori) and to be frequently remissioned after the cure of H. pylori infection. Several previous studies have focused on proliferating lymphocytes but little is known about gastric epithelial change and the duration of the remission after the cure of H. pylori infection. We performed a long-term follow-up investigation on the effects of anti-H. pylori treatment on MALT lymphoma and chronic gastritis at the histologic and molecular levels. Forty-eight patients with low-grade gastric MALT lymphoma and 28 chronic gastritis patients in whom H. pylori infection was eradicated were studied. After eradication, 43 MALT lymphoma patients showed complete histologic remission and continuous remission was observed during follow-up for up to 43 months (mean, 17.8 months). As for epithelial changes after eradication, "emptiness of lamina propria" was more pronounced in the mucosa with MALT lymphoma than that with chronic gastritis, and its severity in MALT lymphoma cases significantly decreased during the observation period whereas the glandular area increased. Cystic change of the fundic gland also occurred more frequently in MALT lymphoma cases than chronic gastritis cases. B-cell clonality before eradication analyzed by reverse transcriptase-polymerase chain reaction (RT-PCR) was detected in almost all MALT lymphoma cases (43cases), but rare in chronic gastritis cases (6cases). After eradication, in spite of histologic regression, 21 MALT lymphoma patients had a persistent monoclonal population during the follow-up period. B-cell monoclonality preceding the malignant transformation was noted in 4 cases. These observations indicate that 1) complete histologic remission of low-grade gastric MALT lymphomas seems stable even if a monoclonal B cell population is detectable in some cases, 2) there may be a stage of disease where monoclonal B cells are present but there is no histologic evidence of MALT lymphoma, and 3) regenerative change of the damaged glands may occur in histologic regressed MALT lymphoma cases. J. Med. Invest. 47:36-46, 2000

Keywords:low-grade MALT lymphoma, epithelial change, empty lamina propria, B-cell clonality, H. pylori, eradication, long-term follow-up

INTRODUCTION
Malignant lymphomas of the mucosa-associated lymphoid tissue (MALT) type are an extra-nodal marginal-zone B-cell lymphoma (1, 2) and the most common of all gastric lymphomas (3). They may be either low or high-grade (2). In low-grade MALT lymphomas of the stomach, small- to medium-sized irregular neoplastic B lymphoid cells, so-called centrocyte-like cells, densely infiltrate the lamina propria and surround, occasionally invading, the reactive-type follicles with germinal centers (1, 4). They also characteristically invade and destroy the epithelium of the gastric glands to form lympho-epithelial lesions (1, 4). High-grade MALT lymphomas, in which neoplastic B lymphoid cells are larger and apparently atypical, may either arise de novo or be transformed from low-grade MALT tumors (2).
It is currently accepted that Helicobacter pylori (H. pylori) plays a causative role not only in chronic active gastritis, peptic ulcer diseases and probably gastric carcinomas (5 -7) but also in gastric low-grade MALT lymphomas (8-10). Clinical studies have reported high remission rates for low-grade MALT lymphomas after the cure of H. pylori infection (8-10). However, little is known about the duration of the remission because long-term follow-up study is limited (11). Moreover, data are also limited with regard to the epithelial changes in the stomach with MALT lymphomas during a long follow-up period.
The development of the polymerase chain reaction (PCR) provides a sensitive tool for the identification of neoplastic B cell populations by the analysis of immunoglobulin heavy chain (IgH) gene. However, the significance of the detection of B-cell monoclonality in gastric lymphoproliferative lesions by PCR re-mains controversial, because monoclonality can be detected not only in neoplastic (8-10) but also in histologically reactive lesions (10, 12, 13). Its clinico-pathological significance has yet to be established.
The aims of this histologic and molecular study on H. pylori-associated gastric lymphoproliferative diseases including low-grade MALT lymphomas and chronic gastritis are 1) to assess the duration of the remission, 2) to clarify the relationship between histologic change and B-cell clonality, and 3) to investigate the epithelial changes over a long-term follow-up period after eradication of H. pylori.

MATERIALS AND METHODS
Patients
Seventy-six patients with H. pylori-infected gastric lymphoproliferative lesions, either chronic gastritis or low-grade MALT lymphoma, were studied. All patients had initially presented with nonspecific upper gastrointestinal symptoms. All previous clini-cal records and histopathologic slides were checked if previous history presented. In MALT lymphoma cases, the involvement of lymph nodes or other organs was assessed by physical examination, routine laboratory tests, chest x-ray, thoracic and abdominal CT and abdominal ultrasound analysis, and the lesion was confined to the stomach.

Tissue specimens
On each occasion, the endoscopy was undertaken by the same endoscopist (Y. U.). The endoscopic findings alone could not discriminate between chronic gastritis and low-grade MALT lymphoma before eradication. Six to ten biopsy specimens were taken from suspicious areas for histology and two others for reverse transcriptase-polymerase chain reaction (RT-PCR) study. The biopsy specimens were immediately fixed with 10% neutral buffered formalin and kept at room temperature for one day. Thin sections were routinely prepared and stained for hematoxylin and eosin and Giemsa. Biopsy materials for RT-PCR were stored without fixation at -8°Cuntil use.

Eradication and follow-up
All 76 patients received anti-H. pylori therapy consisting of a combination of lansoprazole (30mg/day), metronidazole (500mg/day) and amoxicillin(1.5g/day) for 7 to 14 days. All patients underwent endoscopy two to four months after completion of the eradication therapy, and histologic and molecular analysis was done. Subsequently MALT lymphoma patients underwent repeated endoscopic, histologic and molecular analysis at approximately six-month intervals for up to 45 months after anti-H. pylori therapy (mean, 20.2 months) (Fig.1), although 2of them (Cases 26 and 27 in Fig.1) were referred for surgical treatment. Patients whose initial biopsies were assessed as showing a reactive lymphoid infiltrate (chronic gastritis) were followed up only once or twice for up to 33 months (mean, 8.4 months) after eradication treatment.

Histologic criteria
Histopathologic grading for differential diagnosis of gastric lymphoid proliferative lesions was done using the scoring system of Wotherspoon et al. (8). Briefly, grades up to 3 were considered reactive (chronic gastritis) and grades 4 and 5 neoplastic (low-grade MALT lymphoma). The criteria for the diagnosis of low-grade gastric MALT lymphomas were the detection of lymphoepithelial destruction, and the presence of dense diffuse infiltrates of centrocyte-like cells in lamina propria and germinal centers. Presence or absence of lymphoepithelial lesions, lymphoid aggregates and lymphoid follicles was evaluated for differential diagnosis.

Measurement of the degree of "empty lamina propria"
Histologic change in the degree of so-called "empty lamina propria"was morphometrically followed up in 10 selected MALT lymphoma cases for up to 37months. Nine chronic gastritis cases were also studied as a control for up to 22 months. This change was characterized by loose connective tissue with scattered plasma cells, an absence of organized collections of lymphoid cells, and a loss of gastric glands (10). Measurement of the"empty lamina propria" was done as follows:the 3 most prominent areas were selected in each case and photographed at a magnification of 160 times under a microscope. The field of one photograph (110.88cm2) was divided into176 squares and the number of squares containing "empty lamina propria" was counted. The mean value (%, "empty lamina propria" area/total area) was calculated by analyzing 3photographs.

Immunohistochemistry
An immunohistochemical study was done on 20MALT lymphoma and 5chronic gastritis patients. According to previously described methods (14),the procedure was performed with L26 (CD20, 1:100;DAKO, Carpinteria, USA) for B lymphocytes and CD3 (1:100;DAKO) for T lymphocytes.

Reverse transcriptase-polymerase chain reaction (RT-PCR)
RT-PCR for detection of rearranged IgH genes in proliferating lymphoid cells was performed in45 MALT lymphoma and 20 chronic gastritis cases before eradication. Materials in the remaining 3MALT lymphoma cases were not available. After eradication, biopsy materials for RT-PCR were avail-able in all 48 MALT lymphoma cases.
RNA Extraction:Total RNA was extracted from the samples by the modified acid-guanidinium-thiocyanate-phenol-chloroform (AGPC) method (15) using ISOGEN RNA extracting mixture (Nippon Gene, Toyama, Japan) according to the manufacturer's recommendations. When intact 28S and 18S ribo-somal RNAs were visualized on agarose gel, total RNA from the samples was subjected to RT- PCR.
RT-PCR:In order to identify monoclonal B cell proliferation in the gastric lesions, we amplified the entire VH region of IgH cDNA by RT-PCR. For reverse-transcription, 5 μg of total RNA was incubated with M-MLV (Life Technologies, Inc. Gaithersburg, USA) and a dNTP mixture (500μM each;Life Technologies, Inc.) at 42°C for 60 min using a random primer (5μM;Life Technologies, Inc.) in20μl of a reaction mixture. Subsequently, 1μl of the products was subjected to PCR amplification (16). We used a mixture of five leader sequence primers (VH1LS, VH2+4LS, VH3LS, VH5LS and VH6LS;Table1) corresponding to 6 VH regions in IgH genes as an upper-stream primer, and a Cμ sequence (Table1) in the CH1 region as a down-stream primer. We used a pair of primers for GAPDH gene (GAPDH-UP and GAPDH-DN;Table 1) as an internal control for an adequate PCR reaction. PCR was performed as follows;the final concentra-tion of dNTPs and primers in the reaction mixture is 200 μM and 1 μM, respectively. Taq DNA polymerase (Takara, Shiga, Japan) was added to the mixture at a final concentration of 0.05U/μl and the reaction was carried out in Takara Thermal Cycler MP (Takara) under the following conditions;94°Cfor3 min and then 94°C for 1 min, 55°C for 1.5 min, and 72°C for 2.5 min for the desired number of cycles and an extension of 72°C for 4min. In some cases, we performed semi-nested PCR to confirm the reliability and the reproducibility of our experiment using the same upper-stream primer mixture for first PCR and a new down-stream primer (VHFR2;Table1) at the FR2 region in IgH gene. Then, PCR amplification was done under the conditions described above. The PCR products were electro-phoresed in 1.7% agarose gel or 5% polyacrylamide gel and visualized under ultraviolet light after ethidium bromide staining. The size of PCR products was evaluated by comparison with either Pvu2digested pBK-CMV or Hap2 digested pUC19.
We used one B cell lymphoma sample as a positive control and mononuclear cells in peripheral blood prepared from a healthy volunteer as a negative control. To eliminate the possibility of the contami-nation of positive samples in the reagents, one sample without template was also added in each experiment. For each case, the reactions were done at least in duplicate, and only the cases with a dominant band of identically sized fragments were considered to contain a monoclonal B cell population.

Statistical analysis
Statistical differences were evaluated using Stu-dent's t-test and Paired t-test, as applicable. A value of P<0.05 was considered statistically significant.

RESULTS
Histopathology
As defined in Materials and Methods, histologic evaluation of 76 patients distinguished a reactive lymphoid infiltrate (score 2-3) in 28 cases (16males and 12 females) and features suggestive or diagnostic low-grade MALT lymphoma (score 4-5) in 48 (25males and 23 females). Patients with chronic gastritis presented at a mean age of 45.2 years (range 32-55) and low-grade MALT lymphoma at 49.4 years (range 31-71) (Table2). H. pylori organisms were identified in the initial biopsy specimens in all cases by examination of the Giemsa stained sections.
Clinical and histologic comparative results of 76patients are shown in Tables 2, 3 and 4. In cases of chronic gastritis, the mononuclear infiltrate consisted of small lymphocytes with round nuclei and plasma cells, but few or no centrocyte-like cells (Fig.2). As shown in Table 3, the infiltrate was either mild (9 cases), moderate (18 cases) or severe (1 case). Lymphocytes infiltrating epithelium but not lympho-epithelial lesion were present in 9 cases (32%), lymphoid aggregates in 23 (82%) and follicles with germinal centers in 8 (28%). Eradication of H. pylori was achieved in all chronic gastritis cases. These patients were followed up for a mean of 8.4months (range 2-33) after H. pylori treatment. A remarkable decrease in the lymphoid infiltrate was observed at the first post-therapy evaluation (within 2 months);lymphoid aggregates were seen in 9 cases (32%) and lymphoid follicle in one case (3%) (Table 3).
In all 48 MALT lymphoma cases, centrocyte-like cells, lymphoepithelial lesions and lymphoid follicle were present (Fig. 3). Eradication of H. pylori was achieved in all cases (2 after a second treatment course). Patients were followed up for a mean of20.2 months (range3-45) after H. pylori treatment and 17.8 months (range 0-43) after histologic remission. As shown in Fig.1, histologic regression was seen in 43 of 48 patients (89%) and was evident in the first post-treatment biopsy performed 2 to 4 months after eradication in 32 cases (72%) (Fig.4a). Among the remaining 11 cases, histologic regression was observed 4 to 8 months later in 10 cases and 22months later in one case (Case1). Small lymphocyte clusters without centrocyte-like cells were seen in23 cases (48%), and lymphoid follicles in 8 cases(23%) (Table3). In one patient (Case48), relapse of MALT lymphoma was associated with reinfec-tion. This reinfection was retreated and the patient was then in remission stage.
Low-grade MALT lymphoma was also diagnosed by retrospective re-evaluation of histology which was done for biopsy specimens taken from 5 patients 2, 5, 6, 8 and 8 years before the present study, respectively (Cases6, 16, 19, 25 and 27 in Fig.1).Three of these 5 patients'histologic regression occurred after eradication (Fig.1).
Among 48 patients with low-grade MALT lymphoma, 5 patients showed no histologic regression after H. pylori treatment. Two were referred for surgical treatment and histologic analysis revealed a com-ponent of high-grade lymphoma in the resected stomach (Cases 26 and 27 in Fig.1). The remaining 3 patients were followed up for 6, 10 and 13months and showed no histologic change (Cases 23-25 in Fig.1).
We followed up the change in the degree of "empty lamina propria"(Table4). In chronic gastritis cases, there was no significant difference in the degree of empty lamina propria between initial ob-servation (mean, 57%) and final observation (mean, 54%) (P=0.7338). On the other hand, in MALT lymphoma cases, there was a significant difference in the degree of empty lamina propria between initial and final observation (P=0.0002) (Fig.4b). It decreased gradually;in the first observation (4months of post-eradication) the mean value was80%, in 12 months 67%, in 24 months 62% and in28.2 months 59% (Table4).
After eradication, dilatation of the fundic gland was seen in 4 chronic gastritis cases (14%) and 28MALT lymphoma cases (58%). Dilatated glands of various size were surrounded by normal fundic gland and lined by parietal cells, chief cells and mucous secreting cells (Fig.5). Lingular projection of parietal cells was also seen in 4 MALT lymphoma cases.

Immunohistochemistry
Immunohistochemical analysis confirmed the B-cell predominance within the lymphoid infiltrates;the majority were L26 positive-B cells but CD3positive-T cells were also present. Lymphoid ag-gregates and germinal centers which were located deep in the mucosa were of B lineage. Lymphocytes in lymphoepithelial lesions were phenotyped as B lineage cells but a few T cells were also noted.

Reverse transcriptase-polymerase chain reaction (RT-PCR)
Among 20 chronic gastritis cases in which RT-PCR was performed, in 6 there was evidence of a clonal population before H. pylori treatment. Follow-up biopsies after eradication (mean 15.1 months;range 8-33) showed continued absence of histologic fea-tures of lymphoma but RT-PCR showed persistence of a clonal band in 3 of these 6 cases.
As shown in Fig.1, RT-PCR was performed in45MALT lymphoma cases before eradication and a detected monoclonal band in 43 cases and polyclonal band in 2 cases (Cases 13 and 21). After eradication, RT-PCR was performed in 48 MALT lymphoma cases. In the first post-treatment analysis (2 to 4months after eradication), a polyclonal band was detected in 12 cases and monoclonal band in 36cases. During the follow-up period, in 10 of these36 cases there was a change to a polyclonal band while in 26 cases (21 histologically regressed cases and 5 non-regressed cases), the monoclonal band persisted.
In 4 cases (Cases 20 and 32-34 in Fig.1), in spite of histologic features of a reactive lymphoid infiltrate (score2), RT-PCR detected a monoclonal population in the first biopsy. Rebiopsy, done 3 to 19 months later but before eradication, revealed histologic evidence of MALT lymphoma.

DISCUSSION
It is now commonly accepted that gastric B-cell MALT lymphoma is, in most cases, a sequela of H. pylori-induced, T-cell mediated chronic gastritis (17). The finding of regression of MALT lymphoma brought about by eradication of H. pylori infection is an important step in establishing a role for H. pylori in the events leading to MALT lymphoma (8-10). Further evidence of a role for H. pylori came from case reports in which reinfection with H. pylori resulted in a relapse of MALT lymphoma(18, 19).
Epithelial changes of regressed patients with MALT lymphoma during long-term follow-up have not been evaluated in other trials. In the present study, we focused on particularly empty lamina propria that was more pronounced in the case of MALT lymphoma than chronic gastritis. The dem-onstration of a more pronounced empty lamina propria in MALT lymphoma cases indicates possible dense lymphoid infiltration. During the follow-up period (mean 28.2 months;range13-37), the area of empty lamina propria decreased and the glandular area increased which was statistically significant. These results suggest that regenerative change of the damaged glands caused by lymphoepithelial lesions may occur in histologically regressed cases. Reversal of glandular atrophy in patients with chronic gastritis after eradication has been recently reported but the results in these reports have been conflict-ing. While in a few studies reversal of atrophy was found in all or almost all atrophic gastritis patients examined (20-22), only a few or no patients showed reversal of atrophy in other studies (23, 24). This discrepancy may be due to different locations and grades of atrophic gastritis. Reappearance of parietal and chief cells following corticosteroid and anti-ulcer drug sulglycotide treatment in patients with atrophic gastritis has been reported (25-27). These results further support our speculation that regeneration takes place after H. pylori eradication. In the chronic gastritis cases, the gastric gland also recovered but the extent of the recovery was not significant. This may be due to the less dense lymphoid infiltrate in our study. The relatively short-term follow-up (mean 12.2 months;range 12-33) may also explain the discrepancy (20-22).
The presence of dilated glands and intramucosal cysts lined by parietal, mucus and chief cells is the histologic hallmark of fundic gland polyp (28). In our series, there were lesions where histology fulfilled the criteria for the diagnosis of fundic gland polyps but endoscopic features were absent. The fact that cystic change of the fundic gland was seen in 28MALT lymphoma cases (58%) and in 4 chronic gastritis cases (14%) after H. pylori eradication sug-gests a relationship between eradication therapy and cystic change of the fundic gland. Recent reports have found an association between fundic gland polyp and chronic omeprazole use (29, 30). Ang et al. have recently speculated that cysts, parietal cell hypertrophy and lingular projections of parietal cells are likely to be forerunners to the fundic gland polyps in some patients receiving long-term omeprazole therapy (31). However, in our studies we found lingular projection of parietal cells only in4cases. The reason for this discrepancy may be that the treatment period (10-14 days) in our study was shorter than those in the previous study groups.
For the differential diagnosis of reactive and neoplastic lymphoid proliferation in the gastric mucosa, most reports have focused on the detection of monoclonality (8-10). However, the significance of the detection of B-cell monoclonality by PCR remains controversial, because monoclonality is detected not only in neoplastic (8-10) but also in histologically reactive lesions (10, 12, 13). In our series, in fact, 10 cases showed histologic features of a reactive lymphoid infiltrate with a PCR de-tectable monoclonal population. In the absence of histologic signs of low-grade gastric MALT lymphoma, the presence of monoclonal B cells may indicate a special or peculiar form of gastritis with a monoclonal evolution of B cells or suspicious borderline lesion(32). Whether these B cells are prone to evolve into early gastric MALT lymphoma can only be speculated presently. During the follow-up period, histologic features of lymphoma did not develop in the 6 patients eradicated of chronic gastritis with monoclonality in our series and in others (10, 13). In low-grade MALT lymphoma, the cure of H. pylori infection induces complete histologic remission in the majority of cases and the relapse rate is also low. So it is natural that after eradication of H. pylori, histologic features of monoclonal chronic gastritis may be sustained. In contrast, among those 10 monoclonal chronic gastritis cases, histologic evidence of lymphoma was found in 4 cases during follow-up. Before eradication of H. pylori, we per-formed a second histologic examination in these4cases. In addition to our series, MALT lymphoma developed in 14 monoclonal chronic gastritis cases after more than 1 year of follow-up without eradica-tion of H. pylori (33). A PCR detectable population in the absence of histologic features of lymphoma might therefore represent a peculiar tumor pro-gression stage when the clonal population is large enough to be detected by PCR but not by histology. It seems that monoclonality precedes histologic neoplastic transformation and that a diagnosis of B-cell malignancy should not be based solely on the detection of a monoclonal rearrangement of the IgH gene by PCR.
In 21 cases with low-grade MALT lymphoma, there was persistent molecular evidence of a clonal population after histologic regression. It is not clear whether these patients still suffer from the disease or monoclonality only indicates the presence of benign memory B-cell precursors of the malignant clone (32). The patients with monoclonal PCR bands may be more likely to have a relapse of their disease. There was only one case of relapse in our series, and the patient was monoclonal before the relapse was detected. The rapid recurrence of MALT lymphoma in our case could be explained by regression of the pathologic features of MALT lymphoma caused by H. pylori eradication and persistence of the neoplastic cells in a latent form. Subsequently, re-exposure to the H. pylori might result in reactivation of the neoplastic cells and rapid relapse of the MALT lymphoma (18). It is difficult to draw any conclusion at present. Follow-up studies will answer the question of whether patients with monoclonal B cells are at high risk of relapse of the disease.
Low-grade MALT lymphoma was diagnosed by retrospective re-evaluation of histology in 5 cases and a spontaneous cure was not seen in any of these cases within this period. We predict that a natural cure is possible. A decrease in H. pylori in the stomach has been considered to be caused by mucosal atrophy with intestinal metaplasia (34) and mucosal destruction (33). We speculate that intes-tinal metaplasia creates a hostile environment for H. pylori which can lead to a natural cure although it may take a long time.
Most of the studies published to date have not described the duration of the remission of MALT lymphomas after the cure of H. pylori infection. Complete remissions were obtained in 43 of 48patients (89%) in our series and continued through-out the follow-up period (mean 17.8 months;range 0-43), suggesting that most patients who achieve a complete remission do not relapse within a short period of time. Recently, one follow-up study by Neubauer et al. revealed 40 patients in complete remission for a median of14months similar to our analysis (11). Only one relapse of MALT lymphoma was observed in our series but it was associated with reinfection of H. pylori. Case reports on relapse of the low-grade MALT lymphoma are very rare and in most of the cases relapse of MALT lymphoma was associated with reinfection of H. pylori (18, 19).
In 5 patients, no histologic change of low-grade MALT lymphoma was seen after cure of H. pylori infection. Of these 5 patients, 2underwent surgery and a high-grade lymphoma component was detected which was not detected in previous gastric biopsy specimens. Gastric MALT lymphoma may be the result of a multistep process in which a T-cell-driven B-cell growth first leads to monoclonal B-cell prolif-eration (17), and then to the typical lymphoepithelial lesion (8), then a low-grade B-cell clone evolves that may not depend on T-cell help (35) and finally a high-grade B-cell clone can be found which may be dependent upon additional genetic changes such as p53 alteration (36) and bcl-6 protein (37). In a recent review by Isaacson (35), it was pointed out that the regressed low-grade MALT lymphomas following the eradication of H. pylori were all in an early phase. Patients not responding to cure of H. pylori infection may be in a progressive stage of the disease.
In conclusion, our results suggest that 1) regen-erative change of the gastric mucosa occurs in the histologically regressed MALT lymphoma, 2) genetic monoclonality precedes histologic neoplastic trans-formation, and 3) long-lasting remissions can be acquired after the cure of H. pylori infection in patients with early stage low-grade gastric MALT lymphoma. Due to the indolent nature of low-grade MALT lymphoma, the possibility of reinfection and subsequent relapse of MALT lymphoma requires regular follow-up.

ACKNOWLEDGMENTS
We thank Professor Mitsunobu Sato of the Department of Oral and Maxillofacial Surgery, The University of Tokushima School of Dentistry, for his kind cooperation, Ms. Noriko Amo for her tech-nical assistance and Dr. Shingo Wakatsuki and Dr. Hidehisha Horiguchi for statistical assistance.

REFERENCES
1. Isaacson PG, Spencer J, Finn T:Primary B-cell gastric lymphoma. Hum Pathol 17:72-82, 1986
2. Isaacson PG:Gastrointestinal lymphomas of T and B cell types. Mod Pathol 12:151-158, 1999
3. Freeman C, Berg JW, Cutler SJ:Occurrence and prognosis of extranodal lymphomas. Cancer 29:252-260, 1972
4. Isaacson PG, Spencer J:Malignant lymphoma of mucosa associated lymphoid tissue. Histo-pathology 11:445-462, 1987
5. Marshall BJ, Warren JR:Unidentified curved bacilli in the stomach of patients with gastritis and peptic ulceration. Lancet 1(8390):1311-1315, 1984
6. Parsonnet J, Friedman GD, Vandersteen DP, Chang Y, Vogelman JH, Orentreich N, Sibley RK:Helicobacter pylori infection and the risk of gastric carcinoma. N Engl J Med 325:1127-1131, 1991
7. The Eurogast Study Group:An international association between Helicobacter pylori infec-tion and gastric cancer. Lancet 341:1359-1362, 1993
8. Wotherspoon AC, Doglioni C, Diss TC, Pan L, Moschini A, Boni M, Isaacson PG:Regression of primary low-grade B-cell gastric lymphoma of mucosa-associated lymphoid tissue type after eradication of Helicobacter pylori. Lancet 342:575-577, 1993
9. Bayerdorffer E, Neubauer A, Rudolph B, Thiede C, Lehn N, Eidt S, Stolte M:Regression of primary gastric lymphoma of mucosa-associated lymphoid tissue type after cure of Helicobacter pylori infection. Lancet 345:1591-1594, 1995
10. Savio A, Frangin G, Wotherspoon AC, Zamboni G, Negrini R, Buffoli F, Diss TC, Pan L, Isaacson PG:Diagnosis and posttreatment follow-up of Helicobacter pylori positive gastric lymphoma of mucosa-associated lymphoid tis-sue:histology, polymerase chain reaction, or both? Blood 87:1255-1260, 1996
11. Neubauer A, Thiede C, Morgner A, Alpen B, Ritter M, Neabauer B, Wundisch T, Ehninger G, Stolte M, Bayerdorffer E:Cure of Helicobacter pylori infection and duration of remission of low-grade gastric mucosa-associated lymphoid tissue lymphoma. J Natl Cancer Inst 89:1350-1355, 1997
12. Hsi ED, Greenson JK, Singleton TP, Siddiqui J, Schnitzer B, Ross CW:Detection of immuno-globulin heavy chain gene rearrangement by polymerase chain reaction in chronic active gastritis associated with Helicobacter pylori. Hum Pathol 27:290-296, 1996
13. Rudolph B, Bayerdorffer E, Ritter M, Muller S, Thiede C, Neubauer B, Lehn N, Seifert E, Otto P, Hatz R, Stolte M, Neubauer A:Is the polymerase chain reaction or cure of Helicobacter pylori infection of help in the differential diagno-sis of early gastric mucosa-associated lymphatic tissue lymphoma? J Clin Oncol 15:1104-1109, 1997
14. Mollejo M, Lloret E, Menarguez J, Piris MA, Isaacson PG:Lymph node involvement by splenic marginal zone lymphoma : morphological and immunohistochemical features. Am J Surg Pathol21:772-780, 1997
15. Chomozynski P, Sacchi N:Single-step meth-od of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem162:156-159, 1987
16. Nakashiro K, Kawamata H, Hino S, Uchida D, Miwa Y, Hamano H, Omotehara F, Yoshida H, Sato M:Down-regulation of TSC-22 (trarns-forming growth factor beta-stimulated clone22) markedly enhances the growth of a human salivary gland cancer cell line in vitro and in vivo. Cancer Res58:549-555, 1998
17. Hussell T, Isaacson PG, Crabtree JE, Spencer J:The response of cells from low-grade B-cell gastric lymphomas of mucosa-associated lymph-oid tissue to Helicobacter pylori. Lancet 342:571-574, 1993
18. Cammarota G, Montalto M, Tursi A, Vecchio FM, Fedeli G, Gasbarrini G:Helicobacter pylori reinfection and rapid relapse of 27 low-grade B-cell gastric lymphoma. Lancet 345:192, 1995
19. Nobre-Leitao C, Lage P, Cravo M, Cabecadas J, Chaves P, Alberto-Santos A, Correia J, Soares J, Costa-Mira F:Treatment of gastric MALT lymphoma by Helicobacter pylori eradication:A study controlled by endoscopic ultrasono-graphy. Am J Gastroenterol93:732-736, 1998
20. Oberhuber G, Wuendisch T, Rappel S, Stolte M:Significant improvement of atrophy after eradication therapy in atrophic body gastritis. Pathol Res Pract194:609-613, 1998
21. Tucci A, Poli L, Tosetti C, Biasco G, Grigioni W, Varoli O, Mazzoni C, Paparo GF, Stanghellini V, Caletti G:Reversal of fundic atrophy after eradication of Helicobacter pylori. Am J Gas-troenterol 93:1425-1431, 1998
22. Annibale B, Marignani M, Azzoni C, D' Ambra G, Caruana P, D' Adda T, Delle Fave G, Bordi C:Atrophic body gastritis:distinct features associated with Helicobacter pylori infection. Helicobacter 2:57-64, 1997
23. Kuipers EJ, Uyterlinde AM, Pena AS, Roosendaal R, Pals G, Nelis GF, Festen HPM, Meuwissen SUM:Long-term sequelae of Helicobacter pylori gastritis. Lancet 345:1525-1528, 1995
24. Vander Hulst RW, Vander Ende A, Dekker FW, Ten Kate FJ, Weel JF, Keller JJ, Kruizinga SP, Dankert J, Tytgat GN:Effect of Helicobacter pylori eradication on gastritis in relation to cagA:a prospective 1 year follow up study. Gastroenterology 113:25-30, 1997
25. Jeffries GH, Todd JE, Sleisenger MH:The effect of prednisolone on gastric mucosal his-tology, gastric secretion, and vitamin B12absorption in patients with pernicious anemia. J Clin Invest45:803-812, 1966
26. Strickland RG, Fisher JM, Lewin K, Taylor KB:The response to prednisolone in atrophic gastritis:a possible effect on non-intrinsic factor-mediated vitamin B12 absorption. Gut14:13-19, 1973
27. Bazuro GE, Dezi A, Pallotta L, Masci P, Teodori L, Trinca ML, Koch M, Capurso L:Effects of sulglycotide on inflammation and epithelial cell turnover in active Helicobacter pylori+chronic gastritis. A pilot study. Dig Dis Sci 41:22-25, 1996
28. Lee RG, Burt RW:The histopathology of fundic gland polyps of the stomach. Am J Clin Pathol86:498-503, 1986
29. Graham JR:Omeprazole and gastric polyposis in humans. Gastroenterology 104:1584, 1993
30. El-Zimaity HMT, Jackson FW, Graham DY: Fundic gland polyps developing during omeprazole therapy. Am J Gastroenterol 92:1858-1860, 1997
31. Ang ST, Lieberman DA, Ippoliti AF, Weber L, Weinstein WM : Long-term omeprazole therapy in patients with Barrett's esophagus is associ-ated with parietal cell hyperplasia. Gastroen-terology 106:A1016, 1994
32. Thiede C, Morgner A, Alpen B, Wundisch T, Herrmann J, Ritter M, Ehninger G, Stolte M, Bayerdoffer E, Neubauer A:What role does Helicobacter pylori eradication play in gastric MALT and gastric MALT lymphoma? Gastroen-terology 113:S61-S64, 1997
33. Nakamura S, Aoyagi K, Furuse M, Suekane H, Matsumoto T, Yao T, Sakai Y, Fuchigami T, Yamamoto I, Tsuneyoshi M, Fujishima M: B-cell monoclonality precedes the development of gastric MALT lymphoma in Helicobacter pylori-associated chronic gastritis. Am J Pathol 152:1271-1279, 1998
34. Shibata T, Imoto I, Ohuchi Y, Taguchi Y, Takaji S, Ikemura N, Nakao K, Shima T : Helicobacter pylori infection in patients with gastric carcinoma in biopsy and surgical resection specimens. Cancer 77:1044-1049, 1996
35. Isaacson PG:Recent developments in our understanding of gastric lymphomas. Am J Surg Pathol 20:S1-S7, 1996
36. DU MQ, Peng HZ, Singh N, Isaacson PG, Pan LX:The accumulation of p53 abnormality is as-sociated with progression of mucosa-associated lymphoid tissue lymphoma. Blood 86:4587-4593, 1995
37. Omonishi K, Yoshino T, Sakuma I, Kobayashi K, Moriyama M, Akagi T:bcl-6 protein is identified in high-grade but not low-grade mucosa-associated lymphoid tissue lymphomas of the stomach. Mod Pathol11:181-185, 1998

Received for publication November 29, 1999;accepted January 5, 2000.

Address correspondence and reprint requests to Prof. Toshiaki Sano, Department of Pathology, The University of Tokushima School of Medicine, Kuramoto-cho, Tokushima 770-8503, Japan and Fax:+81-88-633-9423.