• marginal zone lymphoma;
  • gastric;
  • extranodal;
  • β2-microglobulin;
  • aerodigestive


  1. Top of page
  2. Abstract


The current study was conducted to determine the incidence of gastric involvement in patients presenting with extranodal marginal zone lymphoma (MZL) outside the gastrointestinal (GI) tract and to identify clinical or laboratory parameters that predict gastric involvement in such cases.


The records of 121 consecutive patients who presented with non-GI extranodal MZL and had undergone esophagogastroduodenoscopy (EGD) as part of their initial workup were retrospectively reviewed. The authors assessed the presence of occult gastric MZL in these patients and possible associations with demographic characteristics; anatomic site of initial presentation; Helicobacter pylori (H. pylori) infection; Zubrod score; International Prognostic Index (IPI); B symptoms; and serum lactate dehydrogenase, hemoglobin, albumin, and β2-microglobulin levels.


The median age at diagnosis of non-GI MZL was 59 years. The most common primary tumor sites were the salivary/parotid gland (32 patients), ocular adnexa (26 patients), skin (19 patients), and lung (17 patients). Twenty-two patients (18%) were found to have gastric involvement on EGD. Using logistic regression analysis, factors found to be associated with gastric involvement included: high IPI score (odds ratio [OR], 3.70; P = .03), female sex (OR, 6.50; P = .02), serum β2-microglobulin level of ≥2.5 mg/L (OR, 3.69; P = .02), and involvement of the aerodigestive mucosal/glandular tissue (OR, 4.50; P = .004). On multivariate logistic analysis, aerodigestive mucosal/glandular sites, H. pylori infection, and an elevated β2-microglobulin level were found to be associated with gastric involvement.


Routine EGD is recommended for patients with non-GI MZL, particularly those with primary aerodigestive mucosal/glandular tissue involvement or those with a high IPI, female sex, elevated serum β2-microglobulin level, or H. pylori infection regardless of the primary tumor site. Cancer 2011. © 2010 American Cancer Society.

Marginal zone lymphomas (MZLs) are a group of non-Hodgkin B-cell lymphomas in the World Health Organization (WHO) classification of neoplastic diseases of hematopoietic and lymphoid tissues.1 Their name is derived from the origin of the lymphoma cells, which is believed to be the B cells normally present in the marginal zones of B-cell follicles.2 MZLs occur as 1 of 3 types: extranodal, lymph node, and splenic. The extranodal type is the most common and was previously known as low-grade B-cell lymphoma of mucosa-associated lymphoid tissue (or MALT lymphoma) in the Revised European-American Lymphoma Classification (REAL).3 Extranodal MZL is a relatively common type of lymphoma, accounting for approximately 70% of all MZLs and 7% to 8% of all non-Hodgkin lymphomas.1, 4, 5

Extranodal MZL usually occurs in regions in which MALT is not present and hence is believed to arise in reaction to stimulation from infectious or autoimmune sources. This stimulation is believed to induce the aggregation of lymphocytes into organized lymphoid tissue.6 The MALT B cells then undergo a pattern of somatic mutations, giving rise to lymphoma cells that replace the population of normal B cells. Approximately 50% of extranodal MZLs occur in the gastrointestinal (GI) tract.2, 7 However, extranodal MZL also occurs outside the GI tract in organs such as the salivary glands, ocular adnexa, lung, thyroid, breast, and liver.8 In a previous study, Raderer et al reported that extragastric MZL was more likely to disseminate than gastric MZL.8 In addition, 2 studies completed at The University of Texas M. D. Anderson Cancer Center (MDACC) demonstrated that approximately 30% of patients presenting with apparently extra-GI MZL had gastric involvement as well, giving rise to the recommendation of routine esophagogastroduodenoscopy (EGD) for all patients with non-GI MZL.9, 10 The purpose of the current study was to provide an update on the incidence of gastric involvement in a large number of patients presenting with apparently non-GI extranodal MZL and to identify clinical or laboratory parameters that predict gastric involvement in such cases. The identification of gastric involvement in those with seemingly isolated non-GI MZL is significant because of the fact that a systemic approach to therapy would be used if simultaneous GI involvement is identified in these patients.


  1. Top of page
  2. Abstract

Approval for this study was obtained from the appropriate Institutional Review Board at MDACC. Patients were identified by searching the institution's Lymphoma Planning Clinic Database, which includes all patients with lymphoma presenting at MDACC between January 1, 1992 and September 30, 2009. Eligibility criteria were 1) a diagnosis of non-GI extranodal MZL; 2) diagnosis at MDACC or referral to MDACC at the time of diagnosis; and 3) completion of EGD performed at diagnosis. Patients were considered to have non-GI extranodal MZL if they presented with non-GI symptoms and if the extranodal MZL was initially diagnosed at an anatomic site other than the GI tract. All biopsy specimens were either obtained at MDACC or, if obtained elsewhere, reviewed by hematopathologists at MDACC. The diagnosis of MZL was confirmed in all cases according to the criteria initially proposed in the REAL classification3 and then incorporated into the 200111 and current 2008 versions of the WHO classification of tumors of hematopoietic and lymphoid tissues.1

Of the 199 patients with non-GI MZL identified at our institution, a total of 121 patients met the inclusion criteria for this study. Clinical and laboratory data extracted from medical records and retrospectively reviewed included demographic information; anatomic site of initial disease presentation; Zubrod score; the presence or absence of B symptoms at the time of presentation; and serum lactate dehydrogenase (LDH), hemoglobin, albumin, and β2-microglobulin levels before therapy. We also assigned a disease stage before EGD according to the Ann Arbor system11 based on findings from computed tomography scans of the neck, thorax, abdomen, and pelvis and from bone marrow aspiration and biopsy, and we used that stage assignment to determine the International Prognostic Index (IPI).12 The presence of gastric MZL and Helicobacter pylori (H. pylori) infection were determined by EGD. Patients who were determined to have multiple sites of involvement were treated with systemic chemotherapy including: 1) doxorubicin, cyclophosphamide, vincristine, and prednisone (CHOP); 2) CHOP plus rituximab; 3) fludarabine, mitoxantrone injection, and dexamethasone; 4) antibiotics; or 5) observation.

Statistical Analysis

Summary statistics were compiled for the study population and chi-square and Fisher exact tests were used to determine whether the distribution of categorical variables differed between patients with and without gastric involvement. Logistic regression analysis was used to examine the influence of clinical and laboratory parameters on the occurrence of gastric involvement.

The odds ratio (OR) was reported. Data analysis was performed using Stata/SE 11.1 statistical software (StataCorp, College Station, Tex).


  1. Top of page
  2. Abstract

Of the 121 identified patients, 41 (34%) were men and 80 (66%) were women. The median age at the time of diagnosis of MZL was 59 years (range, 18 years-85 years). The anatomic sites of disease presentation (which led to the diagnosis of extranodal MZL) were as follows: salivary/parotid gland (32 patients), ocular adnexa (26 patients), skin (19 patients), lung (17 patients), oral cavity (5 patients), thyroid (5 patients), connective tissue (5 patients), nasopharynx (3 patients), breast (3 patients), kidney (2 patients), larynx (1 patient), Waldeyer ring (1 patient), urethra (1 patient), and nasal sinus (1 patient). Twenty-two patients (18%) had evidence of gastric involvement and 24 (20%) had evidence of H. pylori infection on EGD. Characteristics of the 199 patients with extra-GI MZL are shown in Table 1, in which we compared the 121 patients who were the subjects of the current study with the 78 patients who were excluded.

Table 1. Characteristics of All 199 Patients, Comparing the 121 Subjects of the Study With the 78 Who Were Excluded
CharacteristicEGD PerformedEGD Not PerformedP
  1. EGD indicates esophagogastroduodenoscopy; LDH, lactate dehydrogenase; Hgb, hemoglobin; IPI, International Prognostic Index.

Zubrod score   
B symptoms   
Serum LDH level, IU/L   
Serum Hgb level, g/dL   
Serum albumin level, g/dL   
Serum β2-microglobulin level, mg/L
IPI risk category   

Table 2 shows the OR of gastric involvement according to age; sex; Zubrod score; presence or absence of B symptoms; serum LDH, hemoglobin, albumin, and β2-microglobulin levels; H. pylori infection status; disease stage (assigned before EGD); IPI risk category (based on the stage assigned before EGD); and the involvement of aerodigestive mucosal/glandular tissue (ie, parotid gland, submandibular or sublingual gland, lung, oral cavity, nasopharynx, larynx, Waldeyer ring, or nasal sinus). Using logistic regression analysis, gastric involvement was significantly more likely to be associated with female sex (25% vs 5% for men; OR, 6.50 [P = .02]), an elevated serum β2-microglobulin level (≥2.5 mg/L) (26% vs 9% for those with <2.5 mg/L; OR, 3.69 [P = .02]), high or high-intermediate IPI (33% vs 12% for those with low or low-intermediate IPI; OR, 3.70 [P = .03]), and aerodigestive mucosal/glandular presentation (28% vs 8% for patients with nonaerodigestive presentation; OR, 4.5 [P = .004]), specifically parotid gland presentation (OR, 4.2; P = .006). Factors that were found to carry significance on multivariate logistic regression were elevated β2-microglobulin (OR, 4.05; P = .04), gastric H. pylori infection (OR, 5.88; P = .01), and aerodigestive mucosal/glandular involvement at presentation (OR, 4.5; P = .04).

Table 2. Factors Associated With Gastric Involvement
CriteriaPercentage With Gastric InvolvementORP95% CI
  1. OR indicates odds ratio; 95% CI, 95% confidence interval; LDH, lactate dehydrogenase; Hgb, hemoglobin; H. pylori, Helicobacter pylori; IPI, International Prognostic Index.

Age, y
Zubrod score 1.33.590.26-2.13
B symptoms 1.01.990.11-9.18
Serum LDH level, IU/L 2.49.300.44-14.08
 ≥618 (elevated)29%
Serum Hgb level, g/dL
 ≥12 (decreased)13%
Serum β2-microglobulin level, mg/L
Gastric H. pylori infection
IPI risk category 3.70.0301.13-12.06
 High, high-intermediate33%
 Low, low-intermediate12%
Aerodigestive involvement 4.50.0041.11-17.08

Table 3 presents the details of specific sites of extranodal involvement at presentation and the percentage of patients with associated gastric involvement.

Table 3. Gastric Involvement According to Primary Disease Site
Location of Primary DiseaseNo. of Patients With Gastric Involvement
Parotid gland9 of 23 (39%)
Submandibular/sublingual gland1 of 9 (11%)
Ocular adnexa2 of 26 (8%)
Skin3 of 19 (16%)
Lung2 of 17 (12%)
Oral cavity2 of 5 (40%)
Thyroid0 of 5 (0%)
Connective tissue0 of 5 (0%)
Nasopharynx2 of 5 (40%)
Breast0 of 3 (0%)
Kidney0 of 2 (0%)
Larynx0 of 1 (0%)
Waldeyer ring0 of 1 (0%)
Urethra0 of 1 (0%)
Nasal sinus1 of 1 (100%)


  1. Top of page
  2. Abstract

In this review of 121 patients with primary extra-GI MZL, we found that 18% of patients had gastric involvement at the time of diagnosis. We also determined that certain factors were associated with a significantly higher likelihood of gastric involvement: serum β2-microglobulin level ≥2.5 mg/L, high IPI (ie, poor prognosis) according to pre-EGD stage of disease, H. pylori infection, and being female. We also determined that patients with primary MZL located in aerodigestive mucosal/glandular sites, specifically the parotid gland, were more likely to have gastric involvement than those with primary disease located at other sites. Aerodigestive presentation, elevated β2-microglobulin level, and H. pylori infection were 3 factors that were found to remain significant on multivariate logistic regression.

Most of the patients with non-GI MZL were women (66% vs 34% men), and female patients were more likely to have gastric involvement than men (P = .007). To further validate this ratio, we compared the characteristics of all 199 patients who presented with non-GI MZL regardless of whether EGD was performed at diagnosis; the distribution of women to men was still comparable (61% vs 39%). Investigators at MDACC evaluated sex as a prognostic factor in extranodal MZL,13 but that analysis indicated that being female conferred a protective effect on survival. Therefore, the prevalence of gastric involvement in female patients does not necessarily signify shorter overall or recurrence-free survival times. A second factor found to be associated with gastric involvement was an elevated β2-microglobulin level, which has only recently been shown to influence prognosis in patients with non-Hodgkin lymphoma.14-16 In the current study, patients with high serum β2-microglobulin levels (≥2.5 mg/L) were more likely to have simultaneous gastric involvement (P = .016). A third factor we found to be associated with gastric involvement, the IPI, has previously been shown to predict survival in patients with lymphoma.12 IPI is based on 5 criteria (patient age, disease stage, LDH level, Zubrod score, and the presence of >1 extranodal site of disease). We based our IPI on the pre-EGD stage of disease so that the results of the EGD could not be used to increase the IPI risk estimates. The IPI is currently the most widely used prognostic scoring tool in patients with lymphoma.17 Although the majority of the patients in the current study had a low IPI, as might be expected because of the relatively indolent nature of non-GI MZL,18 18 patients (15%) presented with high-intermediate or high-risk disease. Patients with incrementally higher IPI scores had an incrementally increased risk of gastric involvement in addition to the primary disease (P = .03). A fourth factor linked with gastric involvement was the location of the primary disease. Patients with aerodigestive mucosal/glandular tissue involvement were more likely to have simultaneous gastric involvement. This finding may reflect a special homing mechanism of MZL cells within the aerodigestive tract that predisposes them to spread to other regions within the aerodigestive tract.8, 19 Finally, as previously shown, we confirmed the significant association between H. pylori infection and the risk of developing gastric MZL.20

Limitations of the current study include its retrospective nature and the fact that not all patients with extranodal MZL treated at MDACC since 1992 underwent EGD. Of the 199 patients identified with non-GI MZL, 78 were excluded mainly because ordering EGD was not a common practice and it was highly variable among clinicians. However, most patients with non-GI MZL at our institution have undergone EGD after a previous, albeit much smaller, study conducted at our institution revealed a significant association between non-GI MZL and gastric involvement.9 Selecting only the cohort of patients who underwent EGD at their physician's discretion can theoretically affect the significance of some of our results. Conversely, and in an attempt to validate the results of the current study, we compared the clinical characteristics of patients with extra-GI MZL who underwent an EGD at the time diagnosis with those of patients who did not, and found them to be comparable.

The findings of the current study are particularly important for patients believed to have localized disease because such patients are typically treated with radiotherapy alone.9, 21-24 Therefore, routine EGD is recommended for patients with non-GI MZL, most particularly those with aerodigestive mucosal/glandular involvement, or for female patients or those with an elevated serum β2-microglobulin level, H. pylori infection, and higher IPI regardless of the primary site of disease.


  1. Top of page
  2. Abstract
  • 1
    Isaacson P, Chott A, Nakamura S. Extranodal marginal zone lymphoma of mucosa-associated lymphoid tissue (MALT lymphoma). In: SwerdlowSH, CampoE, HarrisN, et al, eds. World Health Organization Classification of Tumours of Haematopoietic and Lymphoid Tissues. Lyon, France: IARC Press; 2008: 214-217.
  • 2
    Zucca E, Bertoni F, Stathis A, Cavalli F. Marginal zone lymphomas. Hematol Oncol Clin North Am. 2008; 22: 883-901, viii.
  • 3
    Harris NL, Jaffe ES, Stein H, et al. A revised European-American classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group. Blood. 1994; 84: 1361-1392.
  • 4
    Zucca E, Conconi A, Pedrinis E, et al. Nongastric marginal zone B-cell lymphoma of mucosa-associated lymphoid tissue. Blood. 2003; 101: 2489-2495.
  • 5
    A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin's lymphoma. The Non-Hodgkin's Lymphoma Classification Project. Blood. 1997; 89: 3909-3918.
  • 6
    Wotherspoon AC. Extragastric MALT lymphoma. Gut. 2002; 51: 148-149.
  • 7
    Thieblemont C, Berger F, Dumontet C, et al. Mucosa-associated lymphoid tissue lymphoma is a disseminated disease in one third of 158 patients analyzed. Blood. 2000; 95: 802-806.
  • 8
    Raderer M, Wohrer S, Streubel B, et al. Assessment of disease dissemination in gastric compared with extragastric mucosa-associated lymphoid tissue lymphoma using extensive staging: a single-center experience. J Clin Oncol. 2006; 24: 3136-3141.
  • 9
    Dabaja BS, Ha CS, Wilder RB, et al. Importance of esophagogastroduodenoscopy in the evaluation of non-gastrointestinal mucosa-associated lymphoid tissue lymphoma. Cancer J. 2003; 9: 321-324.
  • 10
    Liao Z, Ha CS, McLaughlin P, et al. Mucosa-associated lymphoid tissue lymphoma with initial supradiaphragmatic presentation: natural history and patterns of disease progression. Int J Radiat Oncol Biol Phys. 2000; 48: 399-403.
  • 11
    Carbone PP, Kaplan HS, Musshoff K, Smithers DW, Tubiana M. Report of the Committee on Hodgkin's Disease Staging Classification. Cancer Res. 1971; 31: 1860-1861.
  • 12
    [no authors listed]. A predictive model for aggressive non-Hodgkin's lymphoma. The International Non-Hodgkin's Lymphoma Prognostic Factors Project. N Engl J Med. 1993; 329: 987-994.
  • 13
    Mazloom A, Medeiros LJ, McLaughlin PW, et al. Marginal zone lymphoma: factors that affect the final outcome: a study of 275 patients. Cancer. 2010; 116: 4291-4298.
  • 14
    Canovas A, Alonso JJ, Barreiro G, Aguirre C. Prognostic factors in follicular lymphoma: the importance of beta-2 microglobulin. Tumori. 2010; 96: 117-121.
  • 15
    Gui W, Wang T, Wang J, et al. An improved prognostic parameter for non-Hodgkin's lymphoma based on the combination of three serum tumor markers. Int J Biol Markers. 2008; 23: 207-213.
  • 16
    Duletic-Nacinovic A, Stifter S, Marijic B, et al. Serum IL-6, IL-8, IL-10 and beta2-microglobulin in association with International Prognostic Index in diffuse large B cell lymphoma. Tumori. 2008; 94: 511-517.
  • 17
    Friedberg JW, Mauch PM, Rimsza LM, Fisher RI. Non-Hodgkin's lymphoma. In: DeVitaVT, LawrenceTS, RosenbergSA, DePinhoRA, WeinbergRA, eds. DeVita, Hellman, and Rosenberg's Cancer: Principles and Practice of Oncology. 8th ed. Philadelphia: Lippincott Williams & Wilkins; 2008: 2098-2142.
  • 18
    Thieblemont C, Bastion Y, Berger F, et al. Mucosa-associated lymphoid tissue gastrointestinal and nongastrointestinal lymphoma behavior: analysis of 108 patients. J Clin Oncol. 1997; 15: 1624-1630.
  • 19
    Brandtzaeg P, Farstad IN, Haraldsen G. Regional specialization in the mucosal immune system: primed cells do not always home along the same track. Immunol Today. 1999; 20: 267-277.
  • 20
    Santacroce L, Cagiano R, Del Prete R, et al. Helicobacter pylori infection and gastric MALTomas: an up-to-date and therapy highlight. Clin Ter. 2008; 159: 457-462.
  • 21
    Zinzani PL, Magagnoli M, Galieni P, et al. Nongastrointestinal low-grade mucosa-associated lymphoid tissue lymphoma: analysis of 75 patients. J Clin Oncol. 1999; 17: 1254.
  • 22
    Tsang RW, Gospodarowicz MK, Pintilie M, et al. Stage I and II MALT lymphoma: results of treatment with radiotherapy. Int J Radiat Oncol Biol Phys. 2001; 50: 1258-1264.
  • 23
    Cahill M, Barnes C, Moriarty P, Daly P, Kennedy S. Ocular adnexal lymphoma-comparison of MALT lymphoma with other histological types. Br J Ophthalmol. 1999; 83: 742-747.
  • 24
    Thieblemont C, Berger F, Coiffier B. Mucosa-associated lymphoid tissue lymphomas. Curr Opin Oncol. 1995; 7: 415-420.