Main

Extranodal marginal zone B-cell lymphoma of mucosa-associated lymphoid tissue (MALT) is a distinct low-grade lymphoma, which is typically localized long term at the site of origin.1 A preexisting chronic inflammation, such as Helicobacter pylori gastritis, Hashimoto's thyroiditis, or Sjogren's syndrome is considered to influence significantly its development.1 The etiological link between gastric MALT lymphoma and H. pylori gastritis has arisen from the fact that the majority of gastric MALT lymphomas (approximately 70% of cases) show complete regression on eradication of H. pylori alone.2 API2-MALT1 fusion, cloned from t(11;18)(q21;q21), is a gene alteration specific to MALT lymphomas.3, 4 API2 is a member of the IAP (inhibitor of apoptosis) gene family, and is essential for the suppression of apoptosis.5 MALT1, a novel gene, may be involved in NF kappa B activation,6 however, its biological function is not fully elucidated. Recent data have shown that API2-MALT1 fusion leads to an increased inhibition of apoptosis, which thereby helps MALT lymphomas to survive.7

MALT lymphomas occur most often in the stomach, and recently, we systematically analyzed 115 gastric cases.8 As cases positive for API2-MALT1 fusion were totally resistant to H. pylori eradication, we divided the MALT lymphomas into three groups: eradication-responsive and API2-MALT1 fusion-negative (Group A), eradication-resistant and fusion-negative (Group B), and eradication-resistant and fusion-positive (Group C).8 Group A tumors showed superficial gastric wall involvement and a less advanced clinical stage. In contrast, those of Group B and C were often negative for H. pylori infection, and frequently showed a highly advanced clinical stage. A purely low-grade histology was noted in Group C tumors. Some of these features have been reported by other investigators as well.9, 10, 11 These findings suggest that Group A tumors are considerably different from those of Groups B or C in their clinicopathological features.

The immunoglobulin heavy chain gene is formed by rearrangement of the variable (VH), diversity (D), and joining (JH) gene segments at the pre-B cell stage, and there are approximately 50 functional VH fragments that are grouped into seven structurally related families on the basis of at least 80% nucleotide sequence homology.12, 13 A biased usage of particular VH segments has been reported in chronic lymphoid leukemia,14, 15 splenic marginal zone lymphoma,16 and MALT lymphomas of the salivary gland17 and thymus.18 Although gastric MALT lymphoma has been studied more intensively than MALT lymphomas in other sites, its VH gene usage has not been well characterized. Several small series have been performed,19, 20, 21, 22 and no unique use of VH genes has been reported. In this study, we analyzed VH gene usage, somatic mutation, and ongoing mutation status in a large series of gastric MALT lymphomas. In addition, we correlated the VH analysis to the three-group categorization of Groups A, B, and C.

Materials and methods

MALT Lymphoma Cases

A total of 132 cases of gastric MALT lymphoma cases were histologically reviewed according to criteria of the WHO classification for malignant lymphoma.1 The tumor cells were immunohistochemically positive for CD20 and BCL2, but negative for CD3, CD5, CD10, CD23, and cyclin D1. The selection of cases was biased to those that showed resistance to H. pylori eradication. In the preliminary study, distinct monoclonal bands were not obtained in 78. Excluded were six cases with double-gene rearrangements, as detected on 8% polyacrylamide gels, and three cases with nonfunctional immunoglobulin genes as detected by sequence analysis. Consequently, 45 cases (43 biopsy and two resection specimens) were included in this study. Appropriate monoclonal bands were obtained in a similar frequency between eradication-responsive group (23/72 cases, 32%) and eradication-resistant group (22/60 cases, 37%). The study was conducted in accordance with the Declaration of Helsinki. Some of the present cases were included in our previous study.8

Testing for H. pylori infection, H. pylori eradication, and assessment of complete tumor regression have been described elsewhere.8 Briefly, all tumor cases were tested for H. pylori infection using two or more tests including a histologic evaluation, tissue culture, a rapid urease test, a urea breath test, and a serological test for anti-H. pylori immunoglobulin G. In all cases, whether positive or negative for the infection, patients were administered a 2-week course of a proton pump inhibitor (lansoprazole) and a combination of antibiotics (amoxicillin, clarithromycin, and/or metronidazole). The eradication of H. pylori was assessed 6 weeks after completing the antibiotic therapy. Endoscopic examinations were repeated every 3 months until the lymphoma showed complete regression or was judged as resistant. Those that failed to show histological regression 9 months after the successful eradication of H. pylori or that had progressed during follow-up were judged to be resistant.

Detection of the API2-MALT1 Fusion Transcript

All MALT lymphoma cases were tested for the presence of the API2-MALT1 fusion transcript by the multiplex reverse transcription (RT)-polymerase chain reaction (PCR) according to a method that we previously reported.23 Briefly, total RNA was extracted from formalin-fixed, paraffin-embedded specimens by proteinase K digestion. RNA was subjected to first-round multiplex one-step RT-PCR, then to second-round nested multiplex PCRs (three in parallel). The final PCR products were stained with ethidium bromide, and run on 8% polyacrylamide gels. The fusion transcripts were confirmed by direct sequencing. RNA samples known to possess API2-MALT1 fusion were used as positive controls. As an internal control for RNA quality, the ubiquitously expressed β-actin mRNA fragment was amplified. Our RT-PCR assay for the API2-MALT1 fusion transcript using paraffin tissues has a high degree of efficacy, and is capable of 94% of the sensitivity and 100% of the specificity obtained with RT-PCR using frozen materials.24

DNA Extraction, VH Gene Amplification, and Subcloning

Genomic DNA was extracted from the formalin-fixed, paraffin-embedded tumor specimens by overnight digestion with proteinase K. A seminested strategy was used for PCR amplification of the VH genes using a consensus primer for conserved framework-2 (FR2A) and consensus primers for the J region (external LJH and internal VLJH).18, 25, 26 These primers have been used most commonly for VH gene analysis of malignant lymphoma. Subcloning of monoclonal PCR products was performed with pGEM T-easy vector (Promega, Madison, WI, USA) using DNA excised from the sharp rearrangement bands. Recombinant clones were randomly picked up and amplified, then those showing the expected insert size were sequenced using an ABI Prism Big Dye Terminator kit (Applied Biosystems, Foster City, CA, USA) on an automatic DNA sequencer. At least 11 clones were sequenced and analyzed in each lymphoma case.

Sequence Analysis

The DNA sequences were aligned to immunoglobulin heavy chain gene sequences from a well-established database, IgBLAST (http://www.ncbi.nlm.nih.gov/igblast/). VH gene sequences deviating more than 2% from that of the corresponding germline gene were defined as mutated.27 For evaluation of ongoing mutation, the following definitions were used: unconfirmed mutation, a substitution mutation observed in only one of the VH gene clones from the same tumor specimen; and confirmed mutation, observed in more than one of the VH clones from the same tumor specimen.27 Only confirmed mutations were considered as evidence of ongoing mutation; the unconfirmed mutations were instead ascribed to a Taq polymerase error.

Statistical Analysis

Statistical evaluation of data from two groups was performed using the χ2 test, Fischer's exact test, and Student's t-test. All analyses were two-tailed. A probability value of P<0.05 for each test was regarded as statistically significant.

Results

Division of 45 MALT Lymphoma Cases into Three Groups Based on H. pylori Responsiveness and Detection of API2-MALT1 Fusion

Of the 45 cases of gastric MALT lymphoma, 36 were positive for H. pylori infection and the remaining nine were negative (Table 1). Twenty-three cases positive for the infection responded to H. pylori eradication therapy by tumor regression, whereas none of negative cases did. The API2-MALT1 fusion transcript was detected in 12 cases, and all these cases were resistant to the eradication therapy. As in our previous study,8 the lymphoma cases were divided into three groups based on their responsiveness to H. pylori eradication and detection of the API2-MALT1 fusion transcript: Group A, eradication-responsive and fusion-negative (n=23); Group B, eradication-resistant and fusion-negative (n=10); and Group C, eradication-resistant and fusion-positive (n=12).

Table 1 VH gene analysis of gastric MALT lymphoma cases

VH Gene Analysis in Gastric MALT Lymphoma

VH gene analysis was performed by sequencing at least 11 VH clones from each lymphoma, and a total of 583 clones were analyzed. Among the Group A tumors, 18 used the VH3 family and five used VH4, and none of the cases used other VH families. Among Group B tumors, VH3 was used in six cases, VH4 in three, and another was used in one case. As for Group C, VH3 was used in eight cases, VH4 in three, and another gene family in one case (Table 1, Supplementary data). Regarding VH fragments, Group A tumors frequently used VH3-23 in nine cases and VH3-30 in five (total 14/23 cases, 61%). In contrast, Groups B and C tumors used variegated VH fragments, and no dominant VH fragments were noted. Compared with Group A tumors, Group B tumors used VH3-23 and VH3-30 fragments less frequently (1/10 cases, 10%), and this difference was statistically significant (P=0.0094, Table 2). Group C tumors used these two VH fragments (2/12 cases, 17%) with significantly lower frequency than did Group A tumors (P=0.017, Table 2). No dominant D or JH fragments were found in any of the three groups (Table 1).

Table 2 Usage of VH fragments in gastric MALT lymphoma cases (Groups A, B, and C)

Group A, B, and C tumors showed germline VH sequence homology of 95.8%±1.07 (mean±s.e.), 93.9%±1.6, and 95.8%±4.3, respectively. There were no statistical differences between the groups. Deviation from the germline VH exceeding 2% (somatic mutation) was detected in 20/23 (87%), 9/10 (90%), and 11/12 (92%) cases of the respective groups, and these cases were judged to be mutated. Differences were not significant between the groups. In ongoing mutation analysis, tumors with a confirmed, unconfirmed, or negative mutation status, respectively detected in two, 15, and six cases of Group A tumors; in one, six, and three cases of Group B; and in 0, nine, and three of Group C. When assessed by strict criteria for a confirmed mutation, ongoing mutation was detected in 3/45 (7%) of our gastric MALT lymphoma cases.

Discussion

Gastric MALT lymphoma is not a uniform disease judging from its clinicopathological and molecular aspects. By grouping gastric MALT lymphomas into three entities (Groups A, B, and C), we have recently shown that Group A tumors (eradication-responsive and API2-MALT1 fusion-negative) present a typical gastric MALT lymphoma subset, showing such features as superficial gastric wall involvement and a less advanced clinical stage.8 In contrast, Group B tumors (eradication-resistant and fusion-negative) and Group C tumors (eradication-resistant and fusion-positive) are frequently negative for H. pylori infection, and are often at an advanced clinical stage.8, 11 A purely low-grade histology without increased large cells is a feature of the latter.8 To characterize further gastric MALT lymphoma, we performed VH gene analysis of 45 cases by extensive subcloning of monoclonal PCR products.

The most important finding of this study is that inflammation-dependent Group A tumors frequently used particular VH fragments (VH3-23 and VH3-30) as compared with inflammation-independent Group B and C tumors. As there are approximately 50 different functional VH fragments in the human genome that can potentially be used,12, 13 the selective usage of the VH fragments detected in Group A tumors strongly suggests that these tumors may be derived from highly restricted B cell subsets. VH3-23 and VH3-30 have been closely associated with autoimmune disease, with the former found in patients with autoimmune thrombocytopenia,28, 29 Graves' disease,30 and Wegener's granulomatosis,31 and the latter in patients with systemic lupus erythematosus,29 autoimmune thrombocytopenia,28 and Kawasaki's disease.32 In contrast, a biased usage of particular VH fragments has been reported in chronic lymphoid leukemia (VH3-21),14, 15 splenic marginal zone lymphoma (VH1-2),16 and salivary MALT lymphoma (VH1-69).17 We have reported recently that VH3-23 and VH3-30 have been frequently used in thymic MALT lymphoma, a tumor closely associated with autoimmune disease.18, 26 Although the antigen specificity of antibodies encoded by VH3-23 and VH3-30 is currently unknown, it is suggested that B cells using these two VH fragments, which may be associated with H. pylori, are preferentially selected for MALT lymphomagenesis in the stomach. It has been controversial as to whether the majority of inflammation-dependent tumors progress to inflammation-independent tumors. Our data showed that the VH usage pattern, which does not change in the course of tumor progression, was significantly different between tumors that were dependent on inflammation and those that were not. This finding suggests that such progression may not commonly occur, and further supports the hypothesis that API2-MALT1 fusion-positive tumors (Group C), which are often negative for H. pylori infection8, 11 and resistant to eradication,8, 10 may arise independently of chronic inflammation.

Another important finding of this study is that ongoing mutation was infrequent in gastric MALT lymphoma (7% of total cases). This is in contrast to previous observations that ongoing mutation in this lymphoma may be common.19, 21, 22 This discrepancy may be explained by a difference in the criteria used for determination of ongoing mutation. In the present study, ongoing mutation was assessed according to strict criteria for a confirmed mutation,27 which is a substitution mutation observed in more than one of the VH clones from the same tumor specimen. In the previous analyses of gastric MALT lymphoma, ongoing mutation has been judged as present when a substitution mutation is observed in at least one of the VH gene clones from the same tumor specimen. When we similarly assessed our cases, the ongoing mutation rate was increased up to 77%. Conversely, we found that ongoing mutation was infrequent when we re-evaluated raw data of some previous reports, and used the criteria for a confirmed mutation. It would be difficult to determine whether an unconfirmed mutation is a Taq error or a true ongoing mutation. How to resolve this problem is expected to become a subject of future investigation. The majority of our gastric MALT lymphomas showed somatic mutation, which is consistent with a derivation from postgerminal center memory B cells.33, 34

In conclusion, we showed that inflammation-dependent gastric MALT lymphomas frequently used particular VH fragments and that this usage was not found in inflammation-independent tumors. This finding suggests that the former tumors may be derived from a highly restricted, probably H. pylori-associated, B cell subset and that tumor progression from the former to the latter may not commonly occur. Ongoing mutation was not frequent when strictly assessed. However, precise standards need to be established as diagnosis of the presence of ongoing mutation is highly dependent on the type of criteria used.