Helicobacter pylori–negative mucosa-associated lymphoid tissue (MALT) lymphoma of the stomach: A clinicopathologic analysis

Abstract

Objectives

Gastric mucosa-associated lymphoid tissue (MALT) lymphoma is historically associated with Helicobacter pylori (HP) infections in more than 80% of patients. However, the incidence of HP-negative MALT lymphoma has been increasing. The clinicopathologic features have not been well studied, and optimal management strategies remain unclear.

Methods

The pathology database was searched for primary gastric MALT lymphomas diagnosed from 2000 to 2017. The clinical data and the slides were reviewed. The cases were divided for analysis into those with a background of chronic gastritis with HP, chronic gastritis without HP, and without either a background of chronic gastritis or HP.

Results

Of 70 gastric MALT lymphoma cases identified, 26 (37% of total) had chronic gastritis and were positive for HP histologically (n = 23) or were HP positive by additional laboratory testing (n = 3). The remaining 44 (63% of total) cases were HP negative by histology. Within the HP-negative cases, 5 (11% of HP-negative cases) showed histologic gastritis while 39 (89% of HP-negative cases) did not have sufficient evidence of gastritis through review of slides (n = 18) or based on available pathology reports (n = 21). The HP-negative cases without gastritis had higher propensities to show a mass lesion on endoscopy compared with HP-positive cases (37.5% vs 11.1%, P = .02) at the initial diagnosis. The immunophenotype and rate of positive B-cell gene rearrangement were not significantly different between the 2 groups. While all HP-positive patients received antibiotics for HP eradication, treatment in the HP-negative group varied among antibiotics, radiation, rituximab, or chemotherapy. Among HP-negative patients with available follow-up, 13 (39%) showed disease recurrence, similar to the recurrence rate in HP-positive patients; however, no individual from either group has died of the disease thus far.

Conclusions

The incidence of HP-negative MALT lymphoma is increasing, and in our practice, it is currently more common than HP-associated MALT lymphomas. The pathophysiology of HP-negative MALT lymphoma without chronic gastritis remains unclear. Follow-up data in our study suggest that the prognosis of these cases is excellent despite varied management modalities.

KEY POINTS.

• Although historically associated with Helicobacter pylori (HP) infections, recent observations have shown an increase in the incidence of HP-negative mucosa-associated lymphoid tissue (MALT) lymphomas.

• The HP-negative MALT lymphomas are a distinct but poorly characterized subgroup of lymphomas, presenting challenges in diagnosis and management.

• Our study indicates that HP-negative MALT lymphomas have unique clinicopathologic features and generally favorable prognosis, and better understanding of this entity will be important for optimal management strategies.

INTRODUCTION

Extranodal marginal zone lymphoma of mucosa-associated lymphoid tissue (MALT) lymphoma is a non-Hodgkin B-cell ­lymphoma that frequently involves the stomach and has been reported to be associated with Helicobacter pylori (HP) infection in most cases.^1-3H pylori is thought to be responsible for driving B-cell proliferation and establishing the mucosa-associated lymphoid tissue in the gastric mucosa, facilitating progression to MALT lymphoma.^4-6 Other cell types, including T cells and macrophages, may also contribute to disease pathogenesis.^7-9 Although MALT lymphomas are initially HP dependent, subsequent events and associated molecular alterations may cause it to become HP independent. Given the high frequency of gastric MALT lymphomas (initially reported to be associated with HP in up to 80%-90% of cases), it is not surprising that eradication of HP typically leads to its resolution.^10-12 Interestingly, more recent observations have seen a gradual increase in the relative incidence of HP-negative MALT lymphomas, and some studies have reported that these constitute approximately 40% of all gastric MALT lymphomas.^13,14

While there are very few studies on HP-negative MALT lymphomas, it appears that this subgroup is heterogenous and consists of different entities, including HP-associated MALT lymphomas that are negative for HP by routine investigations, as well as those that are truly HP negative. These MALT lymphomas present several challenges as the clinicopathologic features of these lymphomas are poorly characterized, and their etiology, optimal treatment strategies, and long-term outcomes remain unclear. The goal of the study was to evaluate HP-negative MALT lymphomas and evaluate their clinicopathologic features, response to treatment, and prognosis.

MATERIALS AND METHODS

We performed a single-center retrospective analysis using the Yale pathology database to search for primary gastric MALT lymphomas diagnosed from 2000 to 2017. The diagnosis of MALT lymphoma was made using accepted criteria and was confirmed on review. The background gastric mucosa was evaluated for gastritis in uninvolved gastric mucosa, either in the same biopsy or other gastric biopsy specimens taken at the same time, and the gastritis was subjectively scored on a 4-tiered scale (0, none; 1, mild; 2, moderate; 3, severe) using the updated Sydney system.¹⁵ The diagnosis of MALT lymphoma was established in each case based on the presence of abnormal monocytoid B cells admixed with populations of other B and T cells, plasma cells, and often other inflammatory cells; presence of a destructive growth pattern with formation of lymphoepithelial lesions; and, importantly, exclusion of other subtypes of non-Hodgkin lymphoma. Cases with any high-grade lymphoma component or diffuse large cell lymphoma (DLBCL) were excluded. Presence of HP in each case was evaluated by immunohistochemistry (IHC), and staining was performed on 1 or more slides that always included slides where the background gastritis was suspicious for HP infection. The clinical data and the pathology slides were reviewed especially with regards to additional methods of HP testing and treatment. H pylori negativity was defined as negative histology (including IHC or special stains). In addition, 16 of 39 HP-negative cases had at least 1 laboratory test for HP (eg, urea breath test, serology, stool antigen, or rapid urease test). The remaining 23 HP-negative cases by IHC had no additional laboratory testing performed for HP. The cases were divided into HP-positive and HP-negative MALT lymphomas, and HP-negative MALT lymphomas were further subdivided into those with and without a background of chronic gastritis FIGURE 1. The clinical features, endoscopic appearances, histologic features, treatment modalities, and outcomes were obtained from the electronic medical record. Data on immunophenotyping and molecular tests were also recorded.

FIGURE 1.

Study design and inclusion criteria. A total of 70 cases of primary gastric mucosa-associated lymphoid tissue (MALT) lymphomas were included in the study with 63% Helicobacter pylori (HP) negative and 37% HP positive. The categorization of cases for analyses in Tables 1 and 2 was based on HP status and presence or absence of gastritis, evaluation of slide histology (when available), or review of pathology reports as described in the “Materials and Methods” and “Results” sections.

Statistical Analysis

Baseline demographic, clinical, and laboratory data of patients were assessed by χ² test of proportions or Fisher exact test for nominal variables and paired t test or Mann-Whitney U test for continuous variables. All statistical analyses were performed using GraphPad Prism (version 9.5.0; GraphPad Software). In all tests, P values less than .05 were considered statistically significant.

RESULTS

Seventy unique patients with primary gastric MALT lymphomas were identified and divided into HP-negative (n = 44, 63% of total) and HP-positive (n = 26, 37% of total) cases. All HP-positive cases had a background of chronic gastritis (n = 26) and were positive for HP either by IHC staining on histology (n = 23) or by additional laboratory testing for HP (n = 3) FIGURE 1. All 44 HP-negative cases were negative for HP by IHC and had pathology reports available. Twenty-nine HP-negative cases also had slides available for review. Five cases showed histologic gastritis through slide review and were designated as “HP-negative, gastritis-positive” in TABLE 1 and FIGURE 1. The remaining 39 HP-negative cases (designated as “HP-negative, no evidence of gastritis”) showed either no histologic gastritis upon review of slides (18 cases) or were gastritis negative/no mention of the presence of gastritis based on pathology reports (21 cases).

TABLE 1.

Demographics and Characteristics of Gastric MALT Lymphoma Cases

Characteristica	HP negative, no gastritis (n = 39)	HP negative, gastritis positive (n = 5)	HP positive, gastritis positive (n = 26)	P valueb
Age, y	60.8 ± 14.1	71.2 ± 8.7	61.6 ± 14.5	.51
Male sex	11 (28.2)	2 (40.0)	10 (38.5)	.35
Etiologic factors
Autoimmune disease	6 (12.8)	0 (0.0)	1 (6.7)	.09
Other infections	1 (3.7)	0 (0.0)	1 (6.7)	.86
Symptoms present	19 (90.0)	5 (100.0)	14 (87.5)	.96
Endoscopic findings
No visible lesion	3 (12.5)	1 (20.0)	1 (5.6)	.67
Any visible lesion	12 (50.0)	4 (80.0)	15 (83.3)	.02
Mass lesion	9 (37.5)	0 (0.0)	2 (11.1)	.02
Gastric location
Proximal only	33 (84.6)	4 (80.0)	12 (60.0)	.06
Distal only	2 (5.1)	0 (0.0)	4 (20.0)	.15
Multiple regions	4 (10.3)	1 (20.0)	4 (20.0)	.27
Immunophenotype
BCL-2	10 (90.9)	1 (100.0)	6 (75.0)	.41
CD43	9 (52.9)	3 (75.0)	5 (45.4)	.87
CD5	2 (6.9)	1 (20.0)	1 (7.1)	.88
Histologic gastritis
Mild	0 (0.0)	4 (80.0)	0 (0.0)	.04
Severe	0 (0.0)	1 (20.0)	26 (100.0)	<.0001
Positive B-cell gene rearrangement	29 (85.3)	4 (100.0)	13 (76.5)	.67
Positive t(11;18)	1 (25.0)	0 (0.0)	0 (0.0)	.20
Therapy
Antibiotic	2 (6.5)	2 (40.0)	26 (100.0)	<.0001
Radiation	17 (54.8)	2 (40.0)	0 (0.0)	<.0001
Rituximab (single agent)	8 (25.8)	0 (0.0)	0 (0.0)	<.0001
Chemotherapy	4 (12.9)	1 (20.0)	0 (0.0)	.16
Recurrence	13 (39.4)	1 (20.0)	10 (38.5)	.75

^aNominal variables presented as number (%) and continuous variables as mean ± SD.

^b P value comparing total gastritis-positive vs gastritis-negative cases.

Patient demographics, such as age and sex, were similar between these cohorts TABLE 1. Clinically, 13% of patients who were HP negative without gastritis (n = 6) had an additional diagnosis of autoimmune disease (ie, system lupus erythematosus, Sjögren syndrome, psoriatic arthritis, mixed connective tissue disease, etc), while only 7% of HP-positive patients (n = 1) had a diagnosis of autoimmune disease (ie, Hashimoto thyroiditis) (P = .09) TABLE 1. Other non-HP infections have been associated with antigen-driven proliferation of MALT lymphomas at various sites.¹⁶ In our study, only 1 patient in the HP-negative group and 1 patient in the HP-positive group had clinical and laboratory evidence of an infection (HIV and chronic hepatitis C virus, respectively) that has previously been associated with gastric MALT lymphomas; however, the findings were not statistically different between the 2 groups TABLE 1. Greater than 85% of patients in each cohort exhibited symptoms prior to the diagnosis of MALT lymphoma TABLE 1. These included abdominal pain, dyspepsia, gastrointestinal (GI) bleed, weight loss, constipation, anemia, and so on but were not different between the 2 groups (data not shown).

Most of all MALT lymphoma cases had either a visible lesion (ie, nodule, thick fold, etc) or mass lesion on endoscopy, and only a few cases (n = 5) were detected as an incidental finding on biopsy specimen. Interestingly, HP-negative MALT lymphomas without gastritis were a unique group associated with an increased incidence of mass lesions (37.5% vs 11.1% for HP positive and gastritis positive, P = .02) TABLE 1. In addition, a nonsignificant trend was observed toward increased localization to more proximal gastric regions (ie, cardia, fundus, body) in HP-negative MALT lymphomas without gastritis (84.6% vs 68.0% for HP positive and gastritis positive, P = .06) TABLE 1. Representative micrographs show the morphology of primary gastric MALT lymphoma FIGURE 2, including cases of HP-negative gastric MALT lymphomas FIGURE 3. In all cases, the lymphoma cells were positive for CD20 and negative for cyclin D1 by IHC when performed (ie, 33 of 70 total cases, including all CD5+ or CD5dim+ cases, were tested for cyclin D1 and were found to be negative). BCL-2 was positive in 17 cases (85% of total cases tested), CD43 positive in 17 cases (77% of total cases tested), and CD5 or dim CD5 positive in 4 cases (8% of total cases tested). The immunophenotype was not significantly different between the 2 groups TABLE 1.

FIGURE 2.

Example of a primary gastric mucosa-associated lymphoid tissue (MALT) lymphoma. A, B, H&E sections show that the abnormal lymphoid infiltrate is composed of small lymphoid cells with irregular nuclei and a moderate amount of clear cytoplasm. C, Immunohistochemistry demonstrates that the cells are predominantly CD20-positive B cells admixed with scattered CD3-positive T cells (D). (A, C, D, x100; B, x200)

FIGURE 3.

Histology of Helicobacter pylori–negative mucosa-associated lymphoid tissue (MALT) lymphoma. A, An example of incidentally detected MALT lymphoma in an antral biopsy specimen. The atypical lymphoid infiltrate is seen in the center of the biopsy fragment, while the adjacent antral mucosa is completely devoid of any gastritis. B, Another example that was biopsied as a fundic gland polyp and shows a dense lymphoid infiltrate in addition to the fundic gland polyp at low magnification. The oxyntic mucosa in this biopsy specimen away from the MALT lymphoma and antral mucosa in other biopsy specimens fails to show any chronic gastritis. C, D, Medium- to high-power views of the atypical infiltrate from these examples showing dense lymphoid infiltrate comprising small lymphocytes with irregular nuclear borders and a moderate amount of clear cytoplasm. The glandular destruction is evident (C) and small lymphoepithelial lesion (arrow) is also seen at the edge (D). (A, B, x100; C, D, x200)

Of the total cases, 55 (34 HP negative without gastritis and 21 HP positive and gastritis positive) had B-cell gene rearrangement studies performed. Most (84% of tested cases) were positive for a clonal B-cell gene rearrangement, and no significant differences were observed between the groups TABLE 1. The t(11;18)(q21;q21) API2::MALT1 translocation was identified in only 1 of 4 HP-negative without gastritis cases tested and 0 of 7 HP-positive and gastritis-positive cases tested. All patients in the HP-positive group received antibiotic therapy for HP eradication, whereas the treatment modalities in the HP-negative group varied between antibiotics, radiation, rituximab, and chemotherapy. The outcomes between the groups were similar, with 39% of HP-negative patients without gastritis (n = 13) and 35% of HP-positive and gastritis-positive patients (n = 11) showing recurrence on follow-up (P = .75). No patients have died of the disease thus far in either group.

We performed additional univariate analysis of clinicopathologic variables and their association with recurrence in HP-negative MALT lymphomas without gastritis TABLE 2. Follow-up data through tissue histologic evaluation were available for 33 of the 39 HP-negative cases without gastritis, of which 13 cases (39% of HP-negative cases with follow-up available) showed histologic evidence of recurrence and 20 cases (61% of HP-negative cases with follow-up available) showed no evidence of recurrence on follow-up. We observed a nonsignificant trend toward a higher likelihood of recurrence in HP-negative cases with a mass lesion compared to cases with no visible lesion identified on endoscopy (odds ratio [OR], 5.02; 95% CI, 0.87-24.96; P = .08) TABLE 2. Additionally, HP-negative cases with more distal gastric involvement (ie, antrum or beyond) had a nonsignificant trend toward a lower likelihood of recurrence compared to cases localized to more proximal gastric regions (OR, 0.23; 95% CI, 0.04-1.20; P = .10) TABLE 2. In terms of therapy modalities, HP-negative patients receiving radiation therapy had a trend toward a lower likelihood of recurrence compared to patients receiving other therapies (OR, 0.23; 95% CI, 0.06-1.02; P = .06) TABLE 2. Otherwise, no distinct associations between HP-negative MALT ­lymphoma recurrence and other clinicopathologic variables, including patient demographics (ie, age or sex), presence or absence of symptoms, immunophenotype, histologic gastritis, or molecular/cytogenetic results, were observed TABLE 2.

TABLE 2.

Clinicopathologic Factors Associated With Recurrence in HP-Negative MALT Lymphoma^a

Variable	Odds ratio (95% CI)	P value
Age, y
<65 (reference)
≥65	1.40 (0.34-4.95)	.74
Sex
Male (reference)
Female	1.82 (0.52-6.35)	.49
Etiologic factors
None identified (reference)
Autoimmune disease	0.63 (0.11-4.08)	.62
Symptoms present
No (reference)
Yes	0.67 (0.03-14.12)	.78
Endoscopic findings
No visible lesion (reference)
Visible lesion	0.64 (0.14-2.89)	.60
Mass lesion	5.02 (0.87-24.96)	.08
Gastric location
Proximal (reference)
Distal involvement	0.23 (0.04-1.20)	.10
Immunophenotype
CD43 negative (reference)
CD43 positive	0.86 (0.15-4.85)	.88
CD5 negative (reference)
CD5 positive	5.25 (0.64-72.16)	.15
Histologic gastritis
None (reference)
Mild or severe	0.28 (0.02-2.67)	.28
B-cell gene rearrangement
Negative (reference)
Positive	0.19 (0.01-1.46)	.13
Therapy
Antibiotic	4.44 (0.45-67.00)	.22
Radiation	0.23 (0.06-1.02)	.06
Rituximab, single agent	2.43 (0.56-10.46)	.28
Chemotherapy	1.33 (0.21-7.45)	.77

^aUnivariate analysis using Fisher exact test.

DISCUSSION

Gastric MALT lymphoma has a strong correlation with HP infection and historically has been associated with 60% to 90% of cases.^11,17,18 While there has been limited evidence in the literature of any recent changes in the overall incidence of all primary gastric lymphomas or gastric DLBCL, the incidence of HP-negative gastric MALT lymphomas is increasing in many centers.^13,14,19 In our practice, similar trends are being observed, with the current study showing that HP-negative MALT lymphomas are more common than HP-positive MALT lymphomas (63% vs 37%, respectively). The underlying reason for this changing trend in gastric MALT lymphomas is unclear but may be related to more widespread use of antibiotics resulting in decreasing prevalence of HP infections in Western countries.^20,21 It is also possible that while the prevalence of HP infection is decreasing, other yet unidentified pathogens or antigens are playing a greater role in gastric pathology.

It is well established that there is a causal association between HP and gastric MALT lymphoma, with most patients achieving complete remission with antibiotic therapy to eradicate the organism.²² In contrast, HP-negative MALT lymphomas are presumed to be HP independent, and current National Comprehensive Cancer Network (NCCN) guidelines recommend radiation as the preferred therapy for early-stage disease with a limited role for antibiotic therapy.²³ Our findings show that HP-negative patients receiving radiation therapy (n = 17) had a lower likelihood of recurrence compared to other modalities TABLE 2. In contrast, only 2 patients in the HP-negative group received antibiotic therapy, so a definitive conclusion on antibiotic efficacy cannot be adequately determined in our small retrospective study. Remarkably, a subset of HP-negative MALT lymphomas still responds to antibiotic therapy and was initially reported from several small studies.^24,25 There are possible reasons as to why HP-negative MALT lymphomas respond to antibiotic therapies. Such cases may still be related to HP infection, especially those with a background of HP-like gastritis where the organisms have either disappeared spontaneously or following antibiotic treatment or missed on histology due to a sampling error, thus representing a false-negative HP test.²⁶ Proton pump inhibitor therapy can influence HP testing, causing false-negative results.²⁷ In our cohort, only 3 HP-negative patients were on proton pump inhibitor therapy around the time of initial MALT lymphoma diagnosis, so this is an unlikely explanation for the higher proportion of HP-negative cases in our study. Another explanation is that the causative agent is another organism that responds to antibiotic therapy similar to HP.²⁸ Other species of Helicobacter (eg, Helicobacter heilmannii) have been occasionally reported to be associated with MALT lymphomas, where the organisms can be detected on histology with commonly used stains or antibodies for IHC but can be missed on serology, which is species specific. Interestingly, based on limited data, even HP-negative MALT lymphomas without any histologic evidence of background gastritis also seem to respond to antibiotic therapy with sustained remissions.^13,29 The explanation for this phenomenon is unclear, and it is certainly possible that these cases develop in response to nonmicrobial etiology where the stimulant/antigen disappears or the disease remits spontaneously and the apparent response to antibiotics is just spurious. Nevertheless, several consensus reports and clinical practice guidelines now recommend an initial trial of antibiotic therapy even in HP-negative MALT lymphoma cases.^30,31

The MALT-associated t(11;18)(q21;q21) translocation has been reported in approximately 20% to 25% of all gastric MALT lymphomas based on several larger studies.^32-35 The frequency of the t(11;18)(q21;q21) translocation is more commonly observed in HP-negative gastric MALT lymphomas, ranging from 38% to 53%, based on several studies.^35,36 In our study, the t(11;18)(q21;q21) translocation was present in only 1 of 11 cases of gastric MALT lymphomas tested, including 1 of 4 HP-negative without gastritis cases and 0 of 7 HP-positive and gastritis-positive cases. However, given the limitations of our retrospective study, the paucity of patients tested for the t(11;18)(q21;q21) translocation precludes adequate comparisons and definitive conclusions. Other genetic alterations or translocations were not analyzed in our retrospective study. Interestingly, overexpression of nuclear BCL10 has also been detected at higher frequencies in HP-negative gastric MALT lymphomas^36,37 and has been associated with both t(11;18)(q21;q21) and t(1;14)(p22;q32) IGH/BCL10, which may be an interesting area for future investigation.

A significant proportion of HP-negative MALT lymphomas do not have a background of chronic gastritis but seem to show similar rates of residual disease/recurrence in comparison to HP-positive MALT lymphomas. While those with a background of gastritis may still be HP associated, the pathophysiology of HP-negative MALT lymphoma without chronic gastritis as discussed earlier remains unclear, and this remains one of the most interesting groups in this study. The symptoms ranged from mild abdominal discomfort to severe epigastric pain or GI bleeding. In most of these patients, a visible lesion or mass was identified on endoscopy, and the histology of the tumor remains typical of MALT lymphoma in the stomach. While all our patients with HP-positive MALT lymphoma received HP eradication therapy (n = 26), only a small minority of patients with HP-negative MALT lymphoma without gastritis received HP eradication therapy (n = 2). Many of the patients with HP-negative MALT lymphoma had no residual tumor and a low recurrence rate on follow-up (n = 13, 39% of HP-negative cases with follow-up performed). The follow-up data in our study suggest that the overall prognosis of these cases is relatively good, with a similar recurrence rate, despite varied management modalities compared to HP-positive MALT lymphoma cases TABLE 1. A single-center retrospective study evaluating 57 patients with gastric MALT lymphoma observed complete remission in 31 (80%) of 39 HP-positive cases compared to 5 (56%) of 9 HP-negative cases following HP eradication therapy.³⁸ All HP-negative patients receiving radiotherapy or chemotherapy, as either first-line initial therapy or second-line following HP eradiation therapy, achieved complete remission.³⁸ A recent meta-analysis evaluating eradication therapy as an initial treatment for patients with HP-negative gastric MALT lymphoma showed a complete remission rate of 29.3% in the overall pooled analyses, which was modified by factors including t(11;18)(q21;q21) translocation status, method in diagnosing HP, and timing of patient follow-up.³⁹ No randomized controlled trials are available comparing different treatment modalities and recurrence in patients with HP-negative gastric MALT lymphoma. Clearly, the HP-negative gastric MALT lymphomas without associated chronic gastritis seem to represent a unique subgroup from a clinical and pathophysiologic perspective and require further studies.

In summary, our study supports the observation of an increasing proportion of HP-negative MALT lymphomas among gastric MALT lymphomas similar to other studies and growing clinical notion. This group comprises a subset with gastritis that may be still associated with HP, where the detection is falsely negative for a variety of reasons, and a second subset that has no gastritis and likely has no association with HP. The pathophysiology of HP-negative gastric MALT lymphomas without any associated gastritis remains an interesting disease that needs further study.

Funding: This work was supported by the National Institutes of Health (NIH) National Heart, Lung, and Blood Institute (NHLBI) T32 award to S.X.G. (T32-HL007974).

This article is available for CME credit. Go to academic.oup.com/ajcp/pages/journal_cme to see the latest articles. Thecomplete catalog of journal CME coursescan be found at store.ascp.org.

Conflict of interest disclosure

The authors have nothing to disclose.