Helicobacter pylori infection in patients with gastric involvement by adult T-cell leukemia/lymphoma

Authors


Abstract

BACKGROUND

Gastrointestinal involvement is seen frequently in patients with adult T-cell leukemia/lymphoma (ATLL). The authors previously showed a relatively low prevalence of Helicobacter pylori infection in individuals with human T-cell lymphotropic virus 1 (HTLV-1) infection, including patients with ATLL; however, the correlation between H. pylori infection and ATLL gastric involvement has not been investigated.

METHODS

The authors studied 71 patients with ATLL. Gastric involvement was confirmed by endoscopy and biopsy. H. pylori infection was detected by serology, rapid urease test, and immunohistochemistry on biopsy samples. The expression of adhesion molecules on ATLL cells or their ligands on the vasculature in gastric mucosa was analyzed immunohistochemically. The expression of mucosal addressin cell adhesion molecule 1 (MAdCAM-1) was detected by reverse transcriptase-polymerase chain reaction (RT-PCR) analysis.

RESULTS

Gastric involvement was detected in 21 patients (30%), including 8 patients with acute clinical subtype ATLL and 13 patients with lymphoma type ATLL. The prevalence of H. pylori infection was 86% (18 of 21 patients) in the patients with gastric involvement but only 38% (19 of 50 patients) in the patients without such involvement (P < 0.001). The expression of lymphocyte function-associated antigen 1 (LFA-1) and its ligand, intercellular adhesion molecule 1 (ICAM-1), was most frequent on ATLL cells infiltrating the stomach and was enhanced substantially on vascular endothelium in H. pylori-infected gastric mucosa. Human mucosal lymphocyte antigen 1 also was expressed on infiltrating ATLL cells in the stomach. The expression of MAdCAM-1 mRNA assessed by RT-PCR also was seen selectively in H. pylori-infected patients.

CONCLUSIONS

ATLL cells infiltrate gastric tissues infected with H. pylori, probably through the interaction of adhesion molecules on these cells and their ligands on the vasculature, i.e., through the LFA-1/ICAM-1 pathway. Cancer 2002;94:1507–16. © 2002 American Cancer Society.

DOI 10.1002/cncr.10367

Adult T-cell leukemia/lymphoma (ATLL) is an aggressive, usually fatal T-cell malignancy caused by human T-cell lymphotropic virus type 1 (HTLV-1).1, 2 ATLL is characterized by a high tendency of leukemic cells to infiltrate various organs.3, 4 Gastrointestinal involvement in patients with ATLL also is common.5, 6 Recently, Sakata et al.6 demonstrated gastric infiltration of ATLL cells in 23 of 76 patients with ATLL (30.3%). Cell adhesion molecules expressed on leukemic cells (homing receptors) and on the endothelium (vascular addressins) are considered to play a key role in ATLL cell infiltration,4 a process similar to that described for normal leukocyte transmigration from circulation into the tissue.7

Helicobacter pylori causes one of the most common chronic infections in humans.8 Although it is a noninvasive pathogen,9 persistent infection causes chronic active gastritis, which predisposes the mucosa to peptic ulceration, and is believed to be linked eventually to gastric carcinoma and primary gastric lymphoma, especially lymphoma of the mucosa-associated lymphoid tissue (MALT) type.10, 11 The inflammatory response in H. pylori-associated gastritis (HAG) is characterized by intense infiltration of granulocytes and lymphocytes.12 Emigration of leukocytes to inflammatory sites is mediated by relevant cell adhesion molecules both on vascular endothelial cells and on leukocytes, as mentioned above.4, 7 In fact, several studies have shown that, among various adhesion molecules, the expression of immunoglobulin superfamily intercellular cell adhesion molecule 1 (ICAM-1) was increased substantially on endothelial cells in inflamed gastric mucosa infected with H. pylori, suggesting a major role of ICAM-1 in HAG.13, 14

With regard to the correlation between H. pylori and HTLV-1 infection, we recently demonstrated a lower prevalence of H. pylori in individuals with HTLV-1 infection compared with HTLV-1 negative control participants,15 although the exact reason for this remains unclear. The seroprevalence of H. pylori in HTLV-1 carriers and in patients with ATLL was 48.9% (43 of 88 patients) and 46.7% (21 of 45 patients), respectively, whereas the seroprevalence was 64.0% (184 of 292 participants) in control participants. However, only a few patients with ATLL cell infiltration into the stomach were included in the study, and the implication of H. pylori infection in gastric involvement remains to be determined. These findings prompted us to investigate the relation between gastric involvement by ATLL and H. pylori infection.

In the current study, we found a higher prevalence of H. pylori infection in patients with gastric infiltration by ATLL cells compared with patients who were without gastric involvement. Furthermore, we assessed the mechanism of gastric tropism of ATLL cells, particularly in association with H. pylori infection, with a special reference to adhesion properties of infiltrating leukemic cells and patterns of expression of adhesion molecules on the endothelium of gastric tissues.

MATERIALS AND METHODS

Patients

A total of 71 patients with ATLL were enrolled in this study. The patients included 47 men and 24 women with a mean age of 59 years (range, 33–82 years). ATLL was diagnosed based on the following criteria5, 6, 16: 1) seropositivity against HTLV-1 tested using a particle agglutination kit (Serodia HTLV-1; Fuji-Rebio Company, Tokyo, Japan) and, when indicated, confirmed by an enzyme immunoassay kit (Eitest-ATL; Eisai Company, Tokyo, Japan); 2) lymphoid neoplasia confirmed histologically and/or hematologically (abnormal lymphoid cells with nuclear irregularity of lobulation and clover-leaf forms); 3) CD3+, CD4+, CD8−, CD25+, and CD45RO+ immunophenotypes of tumor cells analyzed using a FACSscan flow cytometer (Becton Dickinson, Mountain View, CA); and 4) monoclonal integration of HTLV-1 provirus in chromosomal DNA by Southern blot analysis.17 The clinical type of ATLL was classified into four groups (acute, lymphoma, chronic, and smoldering) based on the report by the Lymphoma Study Group.16

Endoscopic Examination for Gastric Involvement by ATLL

All patients underwent upper gastrointestinal endoscopy as a systemic check-up for determining the extent of the disease prior to chemotherapy. The presence of gastric involvement by ATLL was determined by endoscopy and biopsy. Biopsy specimens were obtained from all gastric lesions, including tumors, giant folds, ulcerations, erosions, and polypoid lesions, when present, and were fixed in formalin and embedded in paraffin. The deparaffinized and hydrated sections, which were 5-μm thick, were stained with hematoxylin and eosin for histologic examination. Moreover, tissue sections were stained immunohistochemically with the streptavidin-biotin-peroxidase-complex method (Histofine SAB-PO kit; Nichirei Company, Tokyo, Japan), as described previously,18 for phenotyping of neoplastic lymphoid cells. The following monoclonal antibodies were used: B-cell markers CD20 and CD79a; (Dako, Glostrup, Denmark) and T-cell markers CD3 and CD45RO (Dako). Gastric involvement was diagnosed when infiltrated neoplastic cells with the morphologic characteristics described above were positive for the immunophenotype for T-cells in specimens from gastric lesions.6 In some gastric lesions, the integration band of HTLV-1 proviral DNA was detected, allowing a definitive diagnosis.19 Endoscopic features of gastric involvement were divided into the following four types: mass forming, diffuse infiltrating, superficial spreading, and unclassified, according to the classification by Nakamura et al.20

Detection of H. pylori Infection

In all 71 patients, H. pylori status was assessed by both serologic and biopsy-based tests. Serum anti-H. pylori immunoglobulin G antibodies were measured by enzyme-linked immunosorbent assay with a commercial kit (HEL-p TEST™; AMRAD Company, Melbourne, Australia). The cut-off value was determined according to the protocol provided by the manufacturer. During endoscopy, two biopsy specimens were obtained from both the antral and corporal normal appearing gastric mucosa along the greater curvature in each patient with ATLL as well as biopsies from gastric lesions; one was used for immunohistochemistry using anti-H. pylori polyclonal antibody (Dako), and the other was used for the rapid urease test (CLO test; Delta West Company, Bentley, Australia).

Immunohistochemical Analysis of Adhesion Molecules

To assess the adhesion properties of ATLL cells infiltrating the stomach, we performed immunohistochemistry on frozen sections of gastric lesions from 21 patients with ATLL patients who had gastric involvement, which had been snap frozen in optimum cutting temperature compound (Tissue-Tek; Miles Inc., Elkhart, IN) in a ethanol-dry ice mixture and stored at −80 °C using monoclonal antibodies against lymphocyte function-associated antigen 1 (LFA-1; integrin αLβ2; Dako), macrophage 1 antigen (Mac-1; integrin αM&bhr;2; Nichirei Company), very late antigen 4 (VLA-4; integrin α4β2; Dako), and human mucosal lymphocyte antigen 1 (HML-1; integrin αEβ7; Dako). Endothelially expressed adhesion molecules also were studied in situ by immunohistochemical analysis using biopsy samples that had been snap frozen and stored as described above. They were taken from endoscopically intact antral mucosa from 24 patients with ATLL who had H. pylori infection and from 12 patients with ATLL who were without H. pylori infection. In addition, the expression of endothelial addressins also was assessed in involved areas. The following monoclonal antibodies were used: anti-ICAM-1, antivascular cell adhesion molecule 1 (anti-VCAM-1), and anti-E-selectin antibodies (all from Dako). Moreover, vessels were immunostained with anti-von Willebrand factor monoclonal antibody (Dako). We assessed quantitatively the difference in the extent of expression of endothelial adhesion molecules between patients with ATLL who were with and without H. pylori infection based on the method reported by Hatz et al.13 Briefly, two independent observers who were blinded to the diagnosis and experimental results counted the number of von Willebrand factor positive vessels and then counted the number of ICAM-1 positive or VCAM-1 positive vessels on a section serial to that stained for von Willebrand factor, and the ratio of ICAM-1 positive or VCAM-1 positive vessels to von Willebrand factor positive vessels was calculated. Immunohistochemical analysis using antibodies for ICAM-1, VCAM-1, E-selectin, and von Willebrand factor also was performed using frozen sections from gastric lesions, and the percentages were calculated in the same fashion. For a negative control, either each nonspecific isotype antibody was used instead of primary antibodies, or the primary antibodies were omitted.

Flow Microfluorometry

Cell surface expression of LFA-1 on freshly obtained, circulating ATLL cells was determined in 17 patients with the acute subtype of ATLL without gastric involvement by using FACSscan flow cytometer. Staining and flow cytometric analysis were performed using standard procedures, as described previously.21

Reverse Transcriptase-Polymerase Chain Reaction

We analyzed the expression of mucosal addressin cell adhesion molecule 1 (MAdCAM-1) by reverse transcriptase-polymerase chain reaction (RT-PCR) analyses of gastric mucosa. Two additional biopsy samples were obtained from intact antral mucosa of the 24 patients with H. pylori infection and the 12 patients who were negative for the infection during endoscopy, and these samples were cryopreserved at −80 °C until RNA extraction. Total RNA was extracted using a commercial kit according to the instructions provided by the supplier (ISOGEN; Nippon Gene Company, Toyama, Japan). Equivalent amounts of RNA were monitored by absorption at 260 nm and by monitoring the density of 28S and 18S RNAs detected after electrophoresis. After 1 μg of total RNA was reverse transcribed to complementary DNA, the target sequence of MAdCAM-1 was amplified in 35 cycles, each consisting of 1 minute at 94 °C for denaturation, 1 minute at 60 °C for annealing, and 1 minute at 72 °C for extension, followed by a final extension for 5 minutes at 72 °C with specific primers22 using an RT-PCR kit (Takara Shuzo Company, Otsu, Japan). Two primers designed to nucleotide positions 978–999 (TGC GGT GCT GGG ACT GCT GCT C; sense) and 1344–1364 (TCA GGG AGG GGC TTC AGG TCA; antisense) of human MAdCAM-1 cDNA sequence were used for amplification of a 387-base pair product.22 A 10-μL aliquot of each PCR product was analyzed by electrophoresis on a 2% agarose gel containing ethidium bromide, and the bands were examined under ultraviolet light for the presence of amplified DNA. Glyceraldehyde-3-phosphate dehydrogenase gene transcript was amplified routinely as described previously21 and was used as an internal control of the processed RNA for each preparation. All samples were obtained with informed consent in accordance with the Helsinki Declaration.

Statistical Analysis

Statistical analysis was performed with Fisher exact probability tests, chi-square tests, Student t tests, Wilcoxon signed-rank tests, and Mann–Whitney U tests. A P value of < 0.05 was accepted as statistically significant.

RESULTS

Prevalence of H. pylori in Patients with ATLL with Gastric Involvement

Three patients had discordant results between the rapid urease test and the immunohistochemical examination using anti-H. pylori antibody (one patient showed positive features of H. pylori infection histologically but had a negative rapid urease test, whereas two patients had positive rapid urease tests but negative histology). These three patients were seropositive and were classified as H. pylori positive. Among the 71 patients with ATLL, 27 patients (38%) were classified with acute type ATLL, 25 patients (35%) were classified with lymphoma type ATLL, 9 patients (13%) were classified with chronic type ATLL, and 10 patients (14%) were classified with smoldering type ATLL. The incidence of H. pylori infection in each clinical subtype was 48% (13 of 27 patients), 60% (15 of 25 patients), 44% (4 of 9 patients), and 50% (5 of 10 patients), respectively, indicating no significant difference among the four clinical subtypes. The overall prevalence of H. pylori in patients with ATLL was 52% (37 of 71 patients). Gastric involvement was defined in 21 patients (30%), including 8 patients with acute type ATLL and 13 patients with lymphoma type ATLL. None of the patients with chronic or smoldering type ATLL had gastric involvement. In patients with ATLL who had with gastric involvement, the prevalence of H. pylori was 86% (18 of 21 patients), which was significantly higher (P < 0.001) compared with the patients who were without gastric involvement (19 of 50 patients; 38%). When limited to the two subtypes (acute type ATLL and lymphoma type ATLL), there was an apparent difference in the prevalence with respect to the presence or absence of gastric involvement (18 of 21 patients [86%] vs. 10 of 31 patients [32%]; P < 0.001). Conversely, among 37 patients with ATLL who had H. pylori infection, 18 patients (49%) had gastric involvement, whereas gastric involvement was observed only occasionally in H. pylori negative patients with ATLL (3 of 34 patients; 9%; P < 0.001). Table 1 summarizes various clinical parameters and laboratory data for each patient with ATLL who had gastric involvement. There were no significant differences between patients with and without gastric involvement with regard to background characteristics, including age, gender, numbers of peripheral blood leukocytes and atypical lymphocytes, and serum levels of lactate dehydrogenase and calcium.

Table 1. Clinical Information on Patients with Adult T-Cell Leukemia/Lymphoma
PatientAge (yrs)GenderClinical subtypeH. pylori statusEndoscopic classificationHistology of gastric lesionWBCAty-ly (%)LDH (IU/L)Ca (mg/dL)
  1. WBC: white blood cell counts (/mm3); Aty-ly: atypical lymphocytes; M: male; F: female; NE: not evaluable.

150MAcuteNegativeSuperficial spreading, multipleSmall cell1400403907.7
268MAcuteNegativeUnclassified, multiplePleomorphic cell45004813819.4
352FAcutePositiveMass forming, solitaryLarge cell84005362911.0
442MAcutePositiveMass forming, multipleLarge cell24,300848768.8
564MAcutePositiveDiffuse infiltratingLarge cell11,300784839.0
666MAcutePositiveDiffuse infiltratingLarge cell117,3009620159.1
743MAcutePositiveUnclassified, multiplePleomorphic cell22,0003119027.4
840MAcutePositiveDiffuse infiltratingLarge cell10,4002427849.4
959FLymphomaNegativeSuperficial spreading, solitaryLarge cell400004948.6
1078FLymphomaPositiveMass forming, multipleMedium cell401009498.7
1163FLymphomaPositiveMass forming, solitaryMedium cell480008168.8
1260FLymphomaPositiveMass forming, solitaryMedium cell700039667.8
1340MLymphomaPositiveMass forming, multipleLarge cell567005679.4
1448MLymphomaPositiveMass forming, multipleLarge cell520015378.5
1568MLymphomaPositiveMass forming, multipleMedium cell510004148.7
1667MLymphomaPositiveMass forming, solitaryLarge cell11,90005738.7
1762MLymphomaPositiveMass forming, multipleMedium cell12,80007678.8
1869MLymphomaPositiveMass forming, multipleLarge cell940018598.1
1954MLymphomaPositiveMass forming, multiplePleomorphic cell6500016709.3
2061MLymphomaPositiveDiffuse infiltratingLarge cell42000684NE
2147MLymphomaPositiveMass forming, multipleLarge cell470007739.0

Endoscopically, mass-forming type ATLL was seen in 13 patients (62.9%), diffuse infiltrating type ATLL was seen in 4 patients (19.0%), superficial spreading type ATLL was seen in 2 patients (9.5%), and unclassified ATLL types were seen in 2 patients (9.5%). The prevalence of H. pylori in each endoscopic classification was as follows: 12 of 13 patients (92%) with mass-forming type ATLL, 4 of 4 patients (100%) with diffuse infiltrating type ATLL, 1 of 2 patients (50%) with superficial spreading type ATLL, and 1 of 2 patients (50%) with an unclassified type of ATLL. In patients with lymphoma type ATLL, the mass-forming type was the most common endoscopic appearance (11 of 13 patients), whereas, in patients with acute type ATLL, the endoscopic features were variable. Again, gastric ulceration was detected in 4 patients with H. pylori infection.

With regard to the histology of gastric lesions in patients with ATLL, we classified such lesions into five groups based on previous reports.23 They included 12 large cell type lesions, 4 pleomorphic type lesions, 4 medium cell type lesions, and 1 small cell type lesion. The prevalence of H. pylori in each histologic lesion type was 92% (11 of 12 patients), 75% (3 of 4 patients), 100% (4 of 4 patients), and 0% (0 of 1 patient), respectively. However, there was no significant difference in the prevalence with respect to endoscopic and histologic types.

Cell Adhesion Molecules on ATLL Cells

Intense expression of LFA-1 was observed on ATLL cells infiltrating the stomach, especially in patients with gastric involvement by ATLL and H. pylori infection (Fig. 1, Table 2). Among the adhesion molecules on ATLL cells, immunoreactivity for LFA-1 was seen most frequently, with a positive staining rate of 91% (19 of 21 patients). VLA-4 expression was seen in 6 of 21 patients (29%) with gastric involvement, and HML-1 expression was seen in 15 of 21 patients (71%). Immunoreactivity for Mac-1 was never observed. In the H. pylori negative sample, leukemic cells showed immunopositivity only for LFA-1. Based on flow cytometric analysis, the LFA-1 expression rate was 24.6% ± 13.6% (mean ± standard deviation) on circulating ATLL cells from patients without gastric involvement.

Figure 1.

Expression of lymphocyte function associated antigen 1 (A), very late antigen 4 (B), and human mucosal lymphocyte 1 (C) on adult T-cell leukemia/lymphoma cells infiltrating the stomach.

Table 2. Expression of Adhesion Molecules on Vascular Endothelial and Infiltrating Leukemic Cells in Gastric Involved Lesions
PatientClinical subtypeH. pylori statusEndothelial expressionaExpression on infiltrating cellsb
ICAM-1VCAM-1LFA-1VLA-4HML-1Mac-1
++++++++++++
  • ICAM-1: intercellular adhesion molecule 1, VCAM-1: vascular cell adhesion molecule 1; LFA-1: lymphocyte function associated antigen 1; Mac-1: macrophage 1 antigen; VLA-4: very late antigen 4; HML-1: human mucosal lymphocyte antigen 1.

  • a

    Endothelial expression: the percentages of the ratio ICAM-1 and VCAM-1 positive to von Willebrand factor positive vessels (mean ± standard deviation).

  • b

    Expression on inflitrating cells: ++; intense staining, +; positive staining, −; negative staining (number of patients).

6AcutePositive84.2 ± 19.441.8 ± 19.8231024231006
2AcuteNegative53.0 ± 14.014.0 ± 12.7011002002002
12LymphomaPositive76.6 ± 16.139.3 ± 28.57500482820012
1LymphomaNegative4216010001001001
21Total74.9 ± 16.136.5 ± 23.29102071441160021

Cell Adhesion Molecules on Vascular Endothelial Cells

Immunoreactivity for E-selectin was never observed on the endothelium (data not shown). VCAM-1 was expressed constitutively on the vascular endothelium in the gastric mucosa of both H. pylori negative patients and H. pylori positive patients, although there was a significant difference in the percentages of VCAM-1 positive vessels between these groups of patients (P < 0.01) (Fig. 2). The expression of vascular ICAM-1 was enhanced more in H. pylori positive patients compared with H. pylori negative patients (Fig. 3). Furthermore, there also was a significant difference in the percentages of ICAM-1 positive vessels between the two groups (P < 0.001) (Fig. 2). The percentages of ICAM-1 positive vessels and VCAM-1 positive vessels were almost identical between the two observers. With regard to the expression of endothelial adhesion molecules in areas of ATLL cell infiltration, ICAM-1 and VCAM-1, but not E-selectin, were expressed strongly on vessels within the involved area (Fig. 4A,B). ICAM-1 and VCAM-1 also were expressed on the vasculature in uninvolved areas, although their expression was not enhanced as much (Fig. 4D,E). ICAM-1 also was detected on infiltrating ATLL cells (Fig. 4C). When each nonspecific isotype serum was used or when primary antibodies were omitted, the specimens showed no immunoreactivity (data not shown).

Figure 2.

Expression of intercellular cell adhesion molecule 1 (ICAM-1) on vascular endothelial cells in gastric mucosa from antral biopsy specimens. (A) Minimal expression of ICAM-1 was detected in mucosal samples that were negative for Helicobacter pylori. (B) H. pylori-associated gastritis showed enhanced ICAM-1 expression.

Figure 3.

Intercellular adhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule 1 (VCAM-1) immunoreactivity on mucosal vasculature of antral biopsy specimens showing control mucosa negative for Helicobacter pylori and H. pylori-associated gastritis. Results were expressed as the percentage of von Willebrand factor positive (vWf+) vessels that were immunoreactive for ICAM-1 or VCAM-1 in serial sections.

Figure 4.

(A) Intercellular adhesion molecule 1 (ICAM-1) expression on the vasculature in the area of gastric involvement by adult T-cell leukemia/lymphoma. (B) Vascular cell adhesion molecule 1 (VCAM-1) expression on the endothelium in the same gastric lesions. (C) ICAM-1 also was expressed on ATLL cells infiltrating the gastric mucosa. (D) ICAM-1 expression on the vasculature in an uninvolved area. (E) VCAM-1 expression on the vasculature in an uninvolved area.

Expression of the MAdCAM-1 Gene in Antral Biopsy Specimens

mRNA transcript for MAdCAM-1 by RT-PCR was detected in 13 of 24 H. pylori positive patients but in only 1 of 12 H. pylori negative patients (P < 0.05) (Fig. 5). Expression of the MAdCAM-1 gene also was detected in all 21 samples from gastric lesions.

Figure 5.

Expression of specific mRNA for mucosal addressin cell adhesion molecule 1 (MAdCAM-1) within gastric biopsy samples. Reverse transcriptase-polymerase chain reaction products of MAdCAM-1 mRNA were detected as a 387-base pair band on a 2% agarose gel. Lanes 1 and 2: patients with adult T-cell leukemia/lymphoma (ATLL) who were negative for Helicobacter pylori infection; lanes 3–7: Patients with ATLL who were positive for H. pylori infection. Size marker (lane MM) was a 100-base pair DNA ladder (New England Biolabs, Beverly, MA). Bottom: Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA expression shown as an internal control.

DISCUSSION

In the current study, we found a significantly higher prevalence of H. pylori in patients with ATLL who had gastric involvement and only a relatively low prevalence in patients with ATLL who were without gastric involvement, which is consistent with our previous study.15 Sakata et al.6 reported a relatively high prevalence of gastric ulceration (13.2%) in patients with ATLL, although those authors did not provide a reason for their findings. In our series, gastric ulcers were detected in five patients (7%), all of whom were infected with H. pylori. It is possible that the high prevalence of gastric ulcers in the patients with ATLL reported by Sakata et al. is associated with H. pylori infection, which is a major etiologic factor in peptic ulceration.10 We suggest, therefore, that, once they are infected with H. pylori, patients with ATLL are likely to develop gastric lesions, including leukemic cell infiltration and ulceration.

This is the first study that focuses on the relation between H. pylori infection and secondary gastric lesions in patients T-cell malignancies. Little is known about the association between H. pylori infection and primary T-cell lymphoma of the stomach,23, 24 because such lymphomas rarely occur in the gastrointestinal tract.25 To our knowledge, only two reports to date have documented this association, reporting high incidences of H. pylori infection: 11 of 15 patients (73%) and 4 of 5 patients (80%) with primary gastric T-cell lymphoma.23, 24 However, both of those studies included a limited number of patients and assessed H. pylori infection only by histologic examination, which may underestimate its prevalence. Therefore, H. pylori infection may occur more frequently in patients with primary gastric T-cell lymphoma than reported, indicating the possible role of H. pylori in gastric T-cell malignancy. Further large-scale studies are warranted to resolve this issue.

We also showed a considerably high incidence of gastric involvement by ATLL (21 of 71 patients; 30%), in agreement with previous studies reporting an incidence of approximately 5.0–30.3%.6, 26–28 These data indicate that the stomach is one of the organs that shows a high tendency for ATLL cell infiltration, in addition to the liver, spleen, and skin.4, 5 A critical event in ATLL cell infiltration into these tissues is the adhesion of these cells to the vascular endothelium, which is mediated by the interaction between homing adhesion molecules on ATLL cells and vascular addressins on endothelial cells.4 In fact, our study showed that such adhesion molecules (e.g., LFA-1 and VLA-4) were expressed on ATLL cells infiltrating the stomach. Conversely, the ligands of these adhesion molecules, ICAM-1 and VCAM-1, were expressed highly on vascular endothelial cells within the lesions of gastric involvement, indicating that these interactions play a role in gastric involvement by ATLL. In particular, LFA-1 was expressed most often on ATLL cells infiltrating the stomach. It is interesting to note that the activation of LFA-1 is triggered by macrophage inflammatory protein 1,29 in which the protein level has been reported to be increased in H. pylori-infected gastric mucosa.30, 31 In patients with acute subtype ATLL without gastric involvement, the LFA-1 expression rate on circulating ATLL cells was not higher than 24.6% ± 13.6%. Moreover, we confirmed that the expression of ICAM-1 and VCAM-1 on endothelial cells was enhanced markedly in H. pylori-infected mucosa. Collectively, it is possible that H. pylori infection may enhance ATLL cell adhesion to the vasculature and its subsequent transmigration into gastric tissue through the interaction between relevant adhesion molecules, i.e., the LFA-1/ICAM-1 pathway, on both ATLL cells and endothelial cells.

It is noteworthy that HML-1 was expressed frequently on ATLL cells infiltrating the stomach. In agreement with this finding, Otsuka et al.32 reported that peripheral mononuclear cells from patients with ATLL who had gastrointestinal involvement showed significantly higher expression of HML-1 compared with patients without such involvement. Because HML-1 defines a membrane molecule on intestinal intraepithelial T-cells,33 it is possible that HML-1 acts as tissue specific homing receptor in ATLL cell infiltration into the gastrointestinal tract. In addition to HML-1, α4β7, one of the gut-homing receptors, is expressed by only a small proportion of circulating T cells but is expressed by the majority of T cells in the gut.34 The endothelial ligand for α4β7 is MAdCAM-1,34, 35 and it is believed that the adhesion pathway defined by α4β7/MAdCAM-1 represents a pivotal tissue specific homing mechanism for the gastrointestinal tract.4, 5, 35 In the current study, the mRNA for MAdCAM-1 was detected selectively in H. pylori-infected gastric mucosa by RT-PCR analysis, which is consistent with the immunohistochemical study by Dogan et al.36 Those authors demonstrated strong expression of MAdCAM-1 on the vasculature of patients with H. pylori-associated follicular gastritis and MALT type lymphoma, suggesting that its expression is associated closely with H. pylori infection and the formation of MALT. It is noteworthy that Tanaka et al.5 reported the correlation between α4β7 expression on ATLL cells and the presence of gastrointestinal involvement. These findings suggest that gastric infiltration of ATLL cells may be mediated at least in part by the specific interaction between the expression of α4β7 on ATLL cells and of MAdCAM-1 on the inflamed vasculature of H. pylori-infected gastric tissue.

In conclusion, we demonstrated a higher incidence of H. pylori infection in patients with ATLL who had gastric involvement compared with patients who were without gastric involvement. We also showed that LFA-1 always was expressed on infiltrating ATLL cells in the stomach, and the expression of its ligand, ICAM-1, was enhanced substantially on the vascular endothelium in H. pylori-infected gastric mucosa or within the lesions of gastric involvement, suggesting that the interaction between LFA-1 and ICAM-1 plays a role in ATLL cell gastric involvement. More specifically, adhesion molecules like MAdCAM-1 may be associated with gastric involvement by ATLL, especially in patients with H. pylori infection. These adhesion molecules may be suitable therapeutic targets in the treatment of patients with ATLL who have gastric involvement.

Acknowledgements

The authors thank Drs. K. Inoue, S. Ikeda, K. Yamasaki, F. Takeshima, K. Kitsukawa, Y. Onizuka, H. Furusu, and K. Matsunaga for their assistance in the preparation of this article.

Ancillary