Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Marginal Zone Lymphomas (Extranodal/Malt, Splenic, and Nodal)
The term marginal zone refers to a histologic compartment located at the periphery of lymphoid follicles immediately outside their mantle zone. The marginal zone is especially evident in the spleen, although identical areas have been observed in other lymphoid structures, including mesenteric lymph nodes and mucosa-associated lymphoid tissue (MALT) ( Fig. 83.1 ). Ordinarily, it is composed predominantly of B cells that are slightly larger than mantle zone lymphocytes and strongly positive for surface IgM, but weakly positive for IgD (in contrast to mantle zone cells, which are strongly positive for IgD). Marginal zone B lymphocytes are thought to be involved in fast protective responses against pathogenic encapsulated bacteria that do not trigger classic T cell–dependent humoral immunity. They are also assumed to be the physiologic counterpart of a group of indolent non-Hodgkin lymphomas (NHLs) that are currently referred to as marginal zone lymphomas (MZLs) .
NORMAL MARGINAL ZONE CELLS.
(A) Marginal zone cells are seen most readily in sections from the normal spleen. The splenic white pulp typically has three distinctive layers: the germinal center, the mantle zone, and external to this, the marginal zone (see asterisk ). (B) Marginal zone cells are not usually seen in lymph nodes, but for some reason are sometimes present in mesenteric lymph nodes (see asterisk ). They have a similar appearance to those in the spleen. (C) In the gastrointestinal tract, the lymphoid tissue in Peyer’s patches is believed to have a marginal zone equivalent (see asterisk ). The cells are again external to the mantle zone and are believed to traffic between the epithelium and the lymphoid follicle.
Although MZLs share a common denomination, arising from histologic similarities, new genetic findings and longer follow-up studies have established that the currently recognized subtypes of MZLs, initially described in the Revised European-American Lymphoma (REAL) classification, are different diseases. In the most recent 2016 WHO classification of tumors of the hematopoietic and lymphoid tissues, MZLs comprise three distinct entities: extranodal MZL (ENMZL) of MALT, or MALT lymphoma; splenic MZL (SMZL) with or without villous lymphocytes; and nodal MZL (NMZL) with or without monocytoid B cells. Immunoproliferative small intestinal disease (IPSID), also known as alpha heavy chain disease, Seligmann disease, or Mediterranean lymphoma, has been recognized as a variant of ENMZL.
As in other hematological malignancies, a precursor state for MZLs has been recognized. Clonal B-cell lymphocytosis of marginal zone origin (CBL-MZ) is a condition where clonal B cells appear in the blood before the onset of overt MZL. The circulating B cells have phenotypic features consistent with a marginal zone origin, but there is absence of clinical symptoms, splenomegaly, hepatomegaly, or lymphadenopathy. In a subset of patients (15% to 20%), CBL-MZ will progress to an overt MZL, most often SMZL.
In aggregate, MZLs represent approximately 10% of all NHL. Clinically, they behave indolently and have a prolonged course. Therefore, management strategies share similarity with other low-grade lymphomas, although specific biologic characteristics and particular pathophysiologic mechanisms determine unique therapeutic approaches in some of the subtypes. This chapter will summarize the clinical characteristics and current management strategies for the different types of MZLs.
Initial Evaluation of Marginal Zone Lymphoma
Adequate biopsy of a lymph node or a suspicious mass, obtained by endoscopic or conventional means, is preferred. A review by a hematopathologist with expertise in the field is essential to establish the diagnosis. In patients without easily accessible nodes or masses, computed tomography (CT) or ultrasound-guided core needle biopsy is usually well tolerated and may be adequate for diagnosis. Sufficient tissue must be obtained for proper histologic examination and required immunophenotypic and genetic studies. Hence, fine-needle aspiration (FNA) is not appropriate for diagnosis.
Physical examination with special attention to peripheral lymph nodes and the abdomen to look for splenomegaly should be performed. Initial laboratory evaluation should include a complete blood count with evaluation of a peripheral blood smear and basic biochemical studies, including lactate dehydrogenase (LDH) level, which is an important prognostic factor and a potential indicator of transformation from indolent to aggressive lymphoma. Antiglobulin tests and reticulocyte count may be useful to rule out autoimmune hemolytic anemia. Serum protein electrophoresis and immunofixation may demonstrate a monoclonal immunoglobulin. β 2 -microglobulin levels may have prognostic value as in other indolent lymphomas. Viral hepatitis (B and C) and human immunodeficiency virus (HIV) studies to exclude co-existing infections that affect therapeutic approaches should be obtained.
Depending on the primary site of disease, specific procedures may be indicated. Gastrointestinal MZLs may require repeat staging endoscopies, during which an adequate number of biopsies should be obtained. For gastric ENMZL, the most commonly affected site, the European Gastro-Intestinal Lymphoma Study (EGILS) group recommends a mapping procedure with a minimum of 10 biopsies taken from visible lesions and additional ones from macroscopically normal mucosa; the same procedure should be repeated to assess treatment response. Since endoscopic biopsies are not transmural, endoscopic ultrasound (EUS) is a useful way to assess the depth of involvement, which has prognostic implications. Moreover, EUS can be helpful to assess regional lymph node involvement. For head primaries, such as ocular adnexal ENMZL, appropriate directed examination and imaging studies (CT or magnetic resonance) are indicated. Routine endoscopy is suggested for patients with non-gastric MZLs, especially females with primary involvement of upper airways, salivary glands, or lungs, and in patients with high β 2 -microglobulin levels or Helicobacter pylori (H. pylori) infection regardless of primary MZL site.
Staging of Marginal Zone Lymphoma
Imaging studies of the chest, abdomen, and pelvis, usually CT scans, should be obtained to adequately stage the disease. A positron emission tomography (PET) scan has been increasingly used for staging and evaluation of response to therapy in NHL, but it may be less useful in MZL because up to 60% of these lymphomas may be PET-negative, especially in early-stage gastric ENZML. Some MZL series, however, document a PET-positivity rate of up to 80%, with better detection rates seen for head and neck and bronchial versus gastric or ocular MZL. The use of PET scan is particularly important in patients with possible transformation to high-grade pathology. Also for staging, and occasionally for diagnostic purposes, bone marrow aspirate (with morphology and flow cytometry) and biopsy are usually advocated, even in cases in which the likelihood of systemic disease is low, because any positive result has important implications for therapy. The European Society for Medical Oncology (ESMO) clinical practice guidelines for diagnosis of MZLs mandate bone marrow examination for NMZL and SMZL, while they highly recommend it in ENMZL, especially in the non-gastric type and when only local therapy is planned.
Staging of MZL is similar to that of other lymphomas; the Ann Arbor system, or an adapted version, is used most frequently ( Table 83.1 ). Nonetheless, specific staging systems have been adopted for particular sites. Gastrointestinal MZL is often staged according to a modified Ann Arbor scheme, commonly known as the Lugano staging system, which incorporates indices corresponding to depth of mucosal invasion and proximity of affected lymph nodes to the primary lesion (see Table 83.1 ). Of note, the Lugano system has no stage III, and its stage IIE may refer to lesions that extend by contiguity to adjacent organs and not necessarily to secondary involvement of lymph nodes. Since the dissemination patterns of extranodal lymphomas are essentially different from those of primary nodal lymphomas, the EGILS group has proposed a new staging system (the Paris staging system) that is based on the TNM scheme used for solid tumors (see Table 83.1 ). The International Society for Cutaneous Lymphomas (ISCL) and the cutaneous lymphoma task force of the European Organization of Research and Treatment of Cancer (EORTC) have also suggested a new, TNM-based staging system for cutaneous lymphomas other than mycosis fungoides and Sézary syndrome, which may be used for the cutaneous forms of ENMZL ( Table 83.2 ).
Table 83.1
Staging Systems for Gastrointestinal Lymphomas
| Adapted Ann Arbor System | Lugano System | Paris System | Areas Involved a |
|---|---|---|---|
| IE1 | I 1 | T1 N0 M0 | Mucosa to submucosa |
| IE2 | I 2 | T2 N0 M0 | To muscularis propria or subserosa |
| T3 N0 M0 | To serosa | ||
| IIE | T4 N0 M0 | To adjacent organs | |
| IIE1 | II 1 | T1-4 N1 M0 | Regional lymph nodes b |
| IIE2 | II 2 | T1-4 N2 M0 | Non-regional abdominal lymph nodes |
| IIIE | IV | T1-4 N3 M0 | Extra-abdominal lymph nodes |
| IV | T1-4 N0-3 M1 | Distant organs | |
| B1 | Bone marrow |
a In case of more than one visible lesion synchronously originating in the gastrointestinal tract, select the characteristics of the more advanced lesion.
b Anatomic designation of lymph nodes as regional according to site: (1) stomach: perigastric nodes and those located along the ramifications of the celiac artery (i.e., left gastric artery, common hepatic artery, splenic artery); (2) duodenum: pancreaticoduodenal, pyloric, hepatic, and superior mesenteric nodes; (3) jejunum/ileum: mesenteric nodes and, for the terminal ileum only, the ileocolic as well as the posterior cecal nodes; (4) colorectum: pericolic and perirectal nodes and those located along the ileocolic, right, middle, and left colic, inferior mesenteric, superior rectal, and internal iliac arteries.
Table 83.2
International Society for Cutaneous Lymphoma/European Organization of Research and Treatment of Cancer Staging System for Cutaneous Lymphomas Other Than Mycosis Fungoides and Sézary Syndrome
From Kim YH, Willemze R, Pimpinelli N, et al. TNM classification system for primary cutaneous lymphomas other than mycosis fungoides and Sézary syndrome: a proposal of the International Society for Cutaneous Lymphomas (ISCL) and the Cutaneous Lymphoma Task Force of the European Organization of Research and Treatment of Cancer (EORTC). Blood . 2007;110(2):479–484.
| Parameter a | Description | |
|---|---|---|
| T | ||
| T1 | Solitary skin involvement | |
| T1a | Lesion <5 cm diameter | |
| T1b | Lesion >5 cm diameter | |
| T2 | Regional skin involvement: multiple lesions limited to 1 body region or 2 contiguous body regions b | |
| T2a | All-disease-encompassing in a <15 cm diameter circular area | |
| T2b | All-disease-encompassing in a >15 and <30 cm diameter circular area | |
| T2c | All-disease-encompassing in a >30 cm diameter circular area | |
| T3 | Generalized skin involvement | |
| T3a | Multiple lesions involving 2 noncontiguous body regions | |
| T3b | Multiple lesions involving ≥3 body regions | |
| N | ||
| N0 | No clinical or pathologic lymph node involvement | |
| N1 | Involvement of 1 peripheral lymph node region that drains an area of current or prior skin involvement | |
| N2 | Involvement of 2 or more peripheral lymph node regions or involvement of any lymph node region c that does not drain an area of current or prior skin involvement | |
| N3 | Involvement of central lymph nodes | |
| M | ||
| M0 | No evidence of extracutaneous nonlymph node disease | |
| M1 | Extracutaneous nonlymph node disease present | |
a The ISCL/EORTC proposes to defer any stage groupings of the TNM classification until further information is available to validate specific stage grouping strategies.
b Definition of body regions: Head and neck: inferior border—superior border of clavicles, T1 spinous process. Chest: superior border—superior border of clavicles; inferior border—inferior margin of rib cage; lateral borders—midaxillary lines, glenohumeral joints (inclusive of axillae). Abdomen/genital: superior border—inferior margin of rib cage; inferior border—inguinal folds, anterior perineum; lateral borders—midaxillary lines. Upper back: superior border—T1 spinous process; inferior border—inferior margin of rib cage; lateral borders—midaxillary lines. Lower back/buttocks: superior border—inferior margin of rib cage; inferior border—inferior gluteal fold, anterior perineum (inclusive of perineum); lateral borders—midaxillary lines. Each upper arm: superior borders—glenohumeral joints (exclusive of axillae); inferior borders—ulnar/radial-humeral (elbow) joint. Each lower arm/hand: superior borders—ulnar/radial-humeral (elbow) joint. Each upper leg (thigh): superior borders—inguinal folds, inferior gluteal folds; inferior borders—mid-patellae, midpopliteal fossae. Each lower leg/foot: superior borders—mid-patellae, midpopliteal fossae.
c Definition of lymph node regions: Peripheral sites: antecubital, cervical, supraclavicular, axillary, inguinal-femoral, and popliteal. Central sites: mediastinal, pulmonary hilar, paraaortic, iliac.
Extranodal Marginal Zone Lymphoma of Mucosa-Associated Lymphoid Tissue Type
Epidemiology and Manifestations
ENMZL is the most frequent of the MZL subtypes, accounting for approximately 8% of all NHL and two-thirds of all MZLs. An analysis of Surveillance, Epidemiology, and End Results (SEER) data quotes a yearly incidence rate of 1.59 per 100,000 adults in the United States. The median age of presentation is around 60, with a wide range spanning from the third to the ninth decades, and there is a slight female predominance (55%). In contrast to most other indolent lymphomas, ENMZL frequently presents at a localized stage (≈40% stage I and ≈30% stage II), and the risk for systemic dissemination is low (albeit variable depending on primary location), which has important implications for the choice of initial therapy.
The most commonly affected primary site is the mucosa of the gastrointestinal tract, in particular the stomach (approximately 44% of all ENMZL cases, with disease being usually multifocal) followed by the small intestine (≈7%). Ocular structures are also frequently involved (≈12%), namely the orbit (≈40% of all ocular adnexal ENMZL), the conjunctiva (35% to 40%), the lacrimal glands (10% to 15%), and the eyelids (≈10%). Other commonly affected sites include the bronchial mucosa (≈11% of all ENMZL cases), the skin (≈9%), the salivary glands (≈6%), and the thyroid gland (≈6%). Sites that are more rarely reported include Waldeyer pharyngeal lymphoid ring, breast, liver, pancreas, urogenital tract, and central nervous system. Findings at presentation depend on the specific organ affected. Gastric ENMZL may lead to dyspepsia, epigastric pain, nausea, anorexia, and manifestations of gastrointestinal bleeding. Conjunctival ENMZL often forms a painless nodule or plaque that has a “salmon-pink patch” appearance and can be associated with erythema, chemosis, and foreign-body sensation. Primary cutaneous ENMZL frequently presents as multiple red to violaceous papules, plaques, or nodules, most often on the trunk or extremities, in particular the arms, which very uncommonly ulcerate. Salivary and lacrimal gland ENMZL is often preceded by sicca syndrome, with xerostomia or xerophthalmia. B symptoms are uncommon (≈15% of cases). Bone marrow involvement has been reported in 2% to 20% of the cases and is usually more common in non-gastric ENMZL.
IPSID usually affects young adults, with no gender predominance, and is seen most commonly in the Middle East and Northern Africa, usually in low socioeconomic status populations. The disease affects the proximal small bowel diffusely and generally presents with a malabsorption syndrome, with steatorrhea, hypocalcemia, weight loss, abdominal pain, and fever. Cases involving the stomach, the colon, and very rarely the respiratory tract have been described.
Pathobiology and Differential Diagnosis
Etiology
ENMZL is strongly associated with chronic antigenic stimulation, including that deriving from chronic bacterial infections or autoimmune disorders ( Table 83.3 ), although the strength of this correlation for some primary disease sites is discordant among studies, suggesting a possible geographic variation. The common assumption about this association is that continual immune stimulation by bacterial or self-antigens leads to expansion of lymphoid elements in the connective tissue adjacent to the epithelium involved, initially leading to a process of reactive lymphoid hyperplasia. Persistent lymphocytic activation and proliferation predisposes to the accumulation of genetic errors that ultimately may result in antigen-independent growth and, consequently, lymphoma emergence. The histologic distinction between the reactive inflammatory process associated with chronic infection (or autoimmunity) and lymphoma proper may be difficult, in which case demonstration of immunoglobulin gene monoclonality by molecular studies may aid in establishing the diagnosis of lymphoma.
Table 83.3
Chronic Antigenic Stimulation and Extranodal Marginal Zone Lymphoma
| Bacterial Infections | ||
|---|---|---|
| Organism | Site | Prevalence (%) |
| Helicobacter pylori | Stomach | 72–100 |
| Campylobacter jejuni | Intestine (IPSID) | ≈70 |
| Chlamydophila psittaci | Conjunctiva | 0–89 |
| Borrelia burgdorferi | Skin | 0–42 |
| Achromobacter xylosoxidans | Lung | 46 |
| Autoimmune Disorders | ||
|---|---|---|
| Disease | Site | Relative Risk |
| Hashimoto thyroiditis a | Thyroid | 67–80 |
| Sjögren syndrome | Salivary and lacrimal glands | 6.6–30.6 |
a Estimated by assuming cases of thyroid histiocytic lymphoma were ENMZL because these were published before the REAL classification.
Histology
ENMZL is composed predominantly of morphologically heterogenous small B cells. These resemble a spectrum spanning from small lymphocytes with scant cytoplasm to slightly larger cells with nuclei similar to those of centrocytes and having relatively abundant pale cytoplasm (leading possibly to a monocytoid appearance). These cells are located in the outer zone of reactive lymphoid follicles, extend into the interfollicular region, and may sometimes colonize the germinal centers. Larger, immunoblast- or centroblast-like cells may be present in small numbers, but an abundance of these should raise suspicion for diffuse large B-cell lymphoma (DLBCL), which requires different management. According to the most recent WHO classification, the term high-grade MALT lymphoma , denoting the presence of sheets of transformed cells, should not be used and instead these tumors should be diagnosed as DLBCL.
Often there are lymphoid infiltrates invading and destroying glandular structures, with eosinophilic degeneration of epithelial cells (so-called lymphoepithelioid lesions), which are strongly suggestive, albeit not pathognomonic, of progression to lymphoma in cases where the differential diagnosis with reactive lymphoid hyperplasia is in doubt ( Fig. 83.2 ). Plasmacytic differentiation is frequent, especially in association with cutaneous, thyroid, and intestinal (IPSID) ENMZL and may pose differential diagnosis problems with lymphoplasmacytic lymphoma (LPL). These plasmacytoid cells often contain periodic acid–Schiff (PAS)–positive inclusions, known as Dutcher or Russell bodies, depending on whether they are located over the nucleus or in the cytoplasm, respectively. Testing for MYD88 gene mutations, which are detected in the majority of LPLs but rarely found in MZLs, is helpful in these cases.
In gastric ENMZL, histology also plays an important role in establishing the diagnosis of H. pylori infection. All biopsy samples should have sections appropriately stained for its detection (see Fig. 83.2 D). Because proton pump inhibitors (PPIs) may decrease the sensitivity of detection, patients should stop taking these medications for at least 2 weeks before biopsies are obtained.
Immunophenotype
ENMZL cells display common pan B-cell markers, such as CD19 and CD20. They are also usually positive for the complement receptors CD21 and CD35, antigens that are shared with follicular dendritic cells, and also for CD79a. Plasmacytoid cells can be CD138-positive. Helpful in the differential diagnosis with other indolent lymphomas, ENMZL are usually CD5-negative (in contrast to chronic lymphocytic leukemia/small lymphocytic lymphoma [CLL/SLL] and mantle cell lymphoma [MCL]), CD23-negative (in contrast to CLL/SLL), CD10-negative (in contrast to follicular lymphoma [FL]), and cyclin D1-negative (in contrast to MCL). They are also negative for BCL6, which may be helpful to exclude transformation to DLBCL. Except for IPSID, in which tumor cells express a truncated alpha heavy chain without any light chain, most ENMZL are typically positive for IgM or, less commonly, IgA or IgG, with light chain restriction. IgD expression is usually negative or very weak. These immunoglobulins may be secreted, especially when there is significant plasmacytic differentiation, and can give rise to a monoclonal band in the serum protein electrophoresis. The truncated heavy chains of IPSID usually do not appear as a monoclonal band because they co-migrate with other serum proteins but can be detected with anti-alpha heavy chain antibodies on immunofixation. ENMZL can be rarely associated with localized or systemic AL amyloidosis.
Clinical practice guidelines published by ESMO propose that testing for CD20, CD5, CD10, and cyclin D1 by immunohistochemistry (IHC) is mandatory, whereas testing for CD23, IgD, and MYD88 mutation is suggested to aid in the diagnosis of MZLs.
Genetics
Specific chromosomal aberrations have been associated with ENMZL, the frequency of which depends strongly on the primary site of disease ( Table 83.4 ). These abnormalities can be detected by conventional cytogenetics in metaphase plates or through fluorescent in situ hybridization (FISH) of interphase nuclei using specific probes. The most commonly observed abnormality is the t(11;18)(q21;q21), which fuses the BIRC3 (baculoviral inhibitor of apoptosis repeat containing 3, also known as API2, or apoptosis inhibitor-2 protein) and MALT1 (MALT lymphoma translocation-1 protein) genes in chromosomes 11 and 18, respectively, leading to expression of a BIRC3-MALT1 chimeric protein. The native MALT1 is part of a protein complex that includes the BCL10 protein (B-cell lymphoma protein 10) and that indirectly leads to nuclear factor κ-B (NFκB) activation, a process under strict control by several upstream factors. Expression of the fusion protein leads to constitutive activation of NFκB via canonical and noncanonical pathways, which in turn leads to resistance to apoptosis and uncontrolled proliferation. The t(11;18) has a special prognostic significance in gastric ENMZL, because its presence is associated with worse response to antibiotic therapy, which is at least partly due to its higher prevalence in H. pylori –negative gastric ENMZL.
Table 83.4
Frequency of Common Genetic Aberrations in Extranodal Marginal Zone Lymphoma According to Primary Site of Disease
| Genetic Abnormality and Genes Involved | ||||||
|---|---|---|---|---|---|---|
| t(11;18)(q21;q21) | t(14;18)(q32;q21) | t(1;14)(p22;q32) | t(3;14)(p14.1;q32) | +3 a | +18 a | |
| Primary Site | BIRC3/MALT1 | IGH@/MALT1 | IGH@/BCL10 | IGH@/FOXP1 | NFKBIZ, BCL6, FOXP1, … | BCL2, NFATC1, … |
| Lung | 36–53 | 6–10 | 2–7 | 0 | 13–20 | 7 |
| Intestine (non-IPSID) | 13–56 | 0 | 0–13 | 0 | 75 | 13–25 |
| Stomach | 6–26 | 1–5 | 0 | 0 | 11–18 | 6–29 |
| Ocular adnexa | 3–10 | 0–25 | 0 | 0–20 | 30–38 | 14–26 |
| Salivary glands | 0–5 | 0–16 | 0–2 | 0 | 8–55 | 8–19 |
| Skin | 0–4 | 0–14 | 0 | 0–10 | 20 | 4 |
| Thyroid | 0–17 | 0 | 0 | 0–50 | 11–17 | 0–22 |
All values expressed as percentages of cases.
a Mostly partial trisomies. Data summarized based on references noted.
Another, less frequently observed, abnormality is the t(14;18) (q32;q21). This translocation is different from that observed in follicular lymphoma, which involves the BCL2 gene, and instead brings the MALT1 gene under the influence of the immunoglobulin heavy chain gene promoter (IGH@), leading to overexpression of MALT1 and, through mechanisms akin to those of t(11;18), constitutive activation of NFκB. The t(1;14) (p22;q32) is seen even more rarely and causes overexpression of the BCL10 gene, which is placed under control of the IGH@ promoter, and, in turn, activation of the same pathways affected by MALT1. A fourth translocation, t(3;14) (p14.1;q32), described mostly in ocular, cutaneous, and thyroid ENMZL, involves the FOXP1 (forkhead box protein P1) transcription factor and the heavy chain promoter. Although FOXP1 is overexpressed in these tumors, its exact significance in their biology is not fully understood.
Apart from the translocations described, all believed to be mutually exclusive, ENMZL has also been associated with gains of genetic material, in particular partial trisomies of chromosomes 3 (including regions affecting FOXP1, NFκBIZ [NFκB inhibitor zeta], and BCL6 [B-cell lymphoma protein 6]) and 18 (affecting NFATC1 [nuclear factor of activated T cells, cytoplasmic, calcineurin-dependent 1], and BCL2). Gains at 6p25 and losses at 6q (affecting TNFAIP3 [tumor necrosis factor, α-induced protein 3]) and 1p have also been reported.
Consistent with a postgerminal center B-cell origin, ENMZL have rearranged immunoglobulin genes that display somatic hypermutation of their variable regions. In the case of IPSID, there are deletions of the alpha heavy chain gene in the VH and CH1 regions, which result in the production of an abnormal heavy chain that cannot bind light chains to form a complete immunoglobulin molecule.
Therapy for Early-Stage (I/II) Disease
Given its rarity, there are no randomized controlled trials defining the optimal treatment for ENMZL. Often, MZL patients are included in trials that enroll multiple types of indolent NHL. Most recommendations arise from consensus panels based on data from retrospective or uncontrolled prospective trials. The most extensive body of data has been gathered on gastric ENMZL.
Gastric Extranodal Marginal Zone Lymphoma
This form of ENMZL has a strong association with active H. pylori infection. If histologic analysis of the gastric biopsies obtained during staging endoscopy fails to demonstrate H. pylori , noninvasive methods, such as breath tests, stool antigen test, or serology, should be used to exclude the infection. Although not necessarily a marker of active infection, the presence of antibodies against H. pylori in an individual not previously treated for this bacterium implicates it in the pathogenesis of the lymphoma.
All patients with gastric ENMZL should be given H. pylori eradication therapy regardless of stage. The focus of therapy for H. pylori– positive disease is on eradication of the infection with one of the currently recommended regimens for this purpose according to local antibiotic resistance pattern. These commonly combine a 4-week therapy course of a PPI and a 2-week therapy course of clarithromycin plus a second antibiotic, usually amoxicillin or metronidazole (triple therapy). However, this is an issue in flux as resistance to clarithromycin is increasing in several regions. Alternatively, quadruple therapy with a PPI, bismuth, tetracycline, and metronidazole can be used. Eradication of H. pylori should be confirmed with an appropriate test, such as the urea breath test, at least 4 to 6 weeks after finishing antibiotherapy and 2 weeks after discontinuing PPI. Pooled data from several prospective and retrospective studies suggest more than 90% eradication rates after initial antibiotherapy. Persistent infection should be treated with a different course of antibiotics, preferably guided by sensitivity tests, because the same data suggests an eradication rate close to 100% after second- or third-line treatments. Repeat endoscopy with biopsies should be obtained 3 months after completion of antibacterial therapy in order to assess tumor response and also to allow histologic confirmation of H. pylori eradication. If H. pylori is still detected, a different antibiotic combination should be tried and the patient reassessed as above.
Responses to therapy are classified according to biopsy findings on endoscopy. Complete histologic response (complete remission [CR]) should be confirmed by a second endoscopy 3 months later and is managed by observation and regular follow-up thereafter as clinically indicated. The presence of small residual lymphoid aggregates early after H. pylori eradication, corresponding to a category of probable minimal residual disease in the French Study Group of Adult Lymphomas (GELA) grading system for posttreatment evaluation, should also be managed as CR. In some cases, these lymphoid aggregates have been shown to harbor cells with the same monoclonal rearrangements of the original tumor. Nevertheless, these patients do not seem to have an increased risk for relapse and most will have evidence of complete response in a subsequent evaluation.
Patients with overt residual (partial remission [PR]) or stable disease, as long as asymptomatic, can also be managed conservatively, with observation or antibiotherapy as appropriate, for several months. Of interest, responses may occur as late as more than 18 months after completion of antibiotic therapy. If a patient achieves clinical and endoscopic remission after eradication of H. pylori but continues to have persistent microscopic disease on histology, waiting at least 12 months before starting another treatment is reasonable. Progressive or symptomatic disease should be managed with local therapy, with radiotherapy being preferred. Despite its established efficacy in disease control, gastrectomy, because of its immediate morbidity and long-term metabolic complications, is currently reserved for management of rare complications such as perforation or bleeding that cannot be controlled endoscopically.
A meta-analysis of 1436 patients with early stage (IE-IIE1) H. pylori– positive disease on prospective or retrospective studies estimates a CR rate of 78% after eradication of H. pylori , but with individual study remission rates anywhere from 47% to 100%. On univariate analysis of available data from the same studies, adverse risk factors for achieving remission included the presence of t(11;18), stage greater than IE1, proximal (body or fundus) location of lesions, and Western (versus Asian) residence. No good evidence is available to support adjuvant chemotherapy after anti– H. pylori treatment as a means to prevent recurrence in localized gastric MALT lymphomas, although this has only been formally addressed with an additional single agent (chlorambucil).
The management of early-stage H. pylori –negative disease is controversial, with some groups suggesting involved field radiation therapy or, if radiation therapy is contraindicated, systemic therapy with rituximab as initial treatment, whereas others propose a trial of anti- Helicobacter therapy. The rationale for the latter derives from anecdotal reports of CR in H. pylori –negative gastric ENMZL patients that were treated exclusively with antibiotherapy, with a meta-analysis suggesting a response rate of up to 19%. These patients are assumed to have been infected with H. pylori that was missed by diagnostic tests or with a different species of Helicobacter , several of which have been recognized.
Response to involved field radiation therapy (IFRT) is excellent, with some series reporting CR rates of up to 100% with total doses as low as 24 to 30 Gy given over 3 to 4 weeks to the stomach and perigastric nodes. Radiation therapy in early-stage gastric ENMZL has not been associated with increased risk of cardiac death. The lower dose of radiation (24 Gy) was shown to be safe and effective in a phase 3 clinical trial of indolent NHL, including ENMZL. Recurrence rate is very low in these patients, but follow-up is limited for most series reported. International Lymphoma Radiation Oncology Group (ILROG) guidelines provide different volumes of radiation to be used for treatment of each site of ENMZL, with newer definitions of involved-site radiotherapy (ISRT).
The optimal follow-up schedule is unknown, although most recommend periodic endoscopies every 3 to 6 months for the first 5 years and yearly thereafter. Long-term follow-up data in patients with complete remission after H. pylori eradication document a 7.2% relapse rate overall (2.2% per year). Some of these relapses were associated with H. pylori recurrence, and responses to retreatment for H. pylori were seen. Whether most complete remissions equate to cures is a question that will require longer follow-up studies. An additional argument in favor of periodic endoscopies is that H. pylori may be associated with an increased risk for gastric adenocarcinoma.
Molecular studies should not be done routinely in follow-up biopsy samples, outside of a research protocol. Several studies have shown that molecular disease, defined by the presence of residual t(11;18) or monoclonal immunoglobulin as evidenced by polymerase chain reaction (PCR) methods, may still be detected even with complete pathologic remissions. However, this finding is not associated with an increased relapse rate and thus is not currently helpful for clinical management.
Ocular Extranodal Marginal Zone Lymphoma
Localized forms of ocular adnexal ENMZL are most frequently managed with radiation therapy, with treatment specifics dependent on the exact location of the tumor in the orbit. Reported responses are excellent, with CR rates of 83% to 100%. Local recurrence rates vary between 0% and 17%, and distant recurrences can occur in up to 25% over 10 years, although disease-specific survival approaches 100% in most series. The exact site of presentation correlates with the risk for systemic recurrence, the lowest being for conjunctival and the highest for eyelid primaries. Long-term complications related to the use and the dose of radiation therapy, such as cataract formation and xerophthalmia, occur in approximately half the patients. Lower-dose fractionated radiation (4 Gy in two fractions) is associated with 5-year progression-free survival (PFS) of approximately 70% and can be considered in some patients to avoid long-term complications. The fractionation of lower-dose radiotherapy in ocular ENMZL was associated with lower risk of cataract and high durable response rates (96% at 2 years). Despite the growing evidence of the efficacy of lower doses of radiation, the standard recommended dose remains 24 Gy in most patients.
Based on the success of antibiotherapy for H. pylori –associated ENMZL and reports of the presence of Chlamydophila psittaci (by PCR, immunofluorescence, or electron microscopy) in ocular tissues of patients affected by ocular adnexal ENMZL, a course of an antichlamydial antibiotic has been evaluated as initial management for these patients. Some studies have reported PR or CR in a significant fraction of these patients, with a prospective study documenting an overall response rate (ORR) of 65% and suggesting improved response rate and PFS in patients in whom C. psittaci eradication was documented. However, the recommendation to treat with a tetracycline at initial diagnosis is controversial because the reported rates of association with C. psittaci are highly variable—some studies suggest prevalence as high as 80%, whereas others are not able to detect the organism in any of the patients. A meta-analysis has suggested that the benefit of antibiotherapy may be restricted to specific geographic areas and, even then, is likely to be limited.
A recent phase 2 study using intralesional rituximab in 20 patients who received previous therapy for ocular ENMZL showed ORR of 65% with a median follow up of 3.5 years. Intralesional rituximab was well-tolerated and can be considered in patients with ocular ENMZL who relapse or are refractory to initial therapy.
Cutaneous Extranodal Marginal Zone Lymphoma
Results for both surgical excision and radiation therapy for limited disease are comparable, with CR rates approaching 100%. Both approaches have a relapse rate (usually limited to the skin) of around 45%. Encouraging results have also been obtained with intralesional injection of rituximab or α-interferon.
Studies from Europe have suggested an association between Borrelia burgdorferi infection and ENMZL of the skin, although this has not been reproduced in Asian and American studies. In view of this, similar to ocular ENMZL, it has been suggested that a course of a tetracycline may be a reasonable first approach, especially in locations where the association has been documented or when infection by B. burgdorferi is detected.
Immunoproliferative Small Intestinal Disease
Antibiotherapy (varying from single-agent tetracycline to triple-antibiotic therapy with ampicillin, metronidazole, and tetracycline, or an H. pylori regimen) has long been recognized as being able to induce CR in early-stage disease. IPSID has been associated with chronic infection with Campylobacter jejuni . Therefore treatment of early disease is directed at bacterial eradication. Historic rates of remission vary between 30% and 70%, depending on the study. Maximal responses may take more than 5 months of therapy, and relapses occur in a subset of patients.
Other Extranodal Marginal Zone Lymphoma
Early-stage ENMZL in other primary sites is managed in a similar way to ocular adnexal or cutaneous forms. When feasible, surgery can be potentially curative and often may be done with primary diagnostic intent. If full resection is achieved, these patients may be observed. Otherwise, for sites not amenable to surgery or if there is residual disease after surgical excision (i.e., positive margins), radiation therapy is the usual preferred approach. It is important to tailor the radiation amount and site to the clinical setting to avoid irradiation of wide areas of the body, which can potentially result in adverse events. For example, in thyroid ENMZL, the entire thyroid gland should be irradiated. If radiation is contraindicated and the patient is asymptomatic, observation is an option. Otherwise, systemic therapy as for advanced stage, preferably one with minimal toxicity, is appropriate.
Therapy for Advanced-Stage (III/IV) Disease
The data regarding management of advanced-stage disease are also limited, because most large treatment series with long-term follow-up aggregate all indolent lymphomas and include only a small fraction of patients with MZL among other more frequent histologies. Thus most treatment approaches have been modeled after those for FL (see Chapter 82 ), and indeed the most common recommendation is that advanced ENMZL be managed as advanced FL. In any case, there is some suggestion that responses may be better, possibly because of a usually lower burden of disease. Eligible patients should be included whenever possible in clinical trials. It is important to consider that patients with indolent lymphomas, especially MZL, tend to relapse and may do so multiple times. Accordingly, sequencing treatment to avoid toxicity and preserve adequate organ function is necessary to achieve long survival time.
As in other indolent lymphomas, extensive disease is likely incurable with current approaches, which together with a generally slow pace of progression means that systemic treatment is not always indicated. Since no benefit in survival has been demonstrated with early systemic treatment of asymptomatic disease, unless a treatable etiology has been identified, initial management of asymptomatic patients with expectant observation is acceptable. While on this watch-and-wait approach, patients should be reassessed approximately every 3 months with history, physical examination, complete blood counts, basic chemistry, and LDH. Any new symptoms or findings suggestive of transformation should be investigated with a repeat biopsy to rule it out. Routine repeat imaging studies are controversial and should be considered on a case-by-case basis.
Indications for initiating systemic treatment include symptomatic disease due to mass effect or effusion, risk for local compressive disease, bulky lymphadenopathy, symptomatic splenomegaly, B symptoms, cytopenias due to bone marrow involvement, or rapid disease progression. Although unlikely to contribute to regression of established advanced disease, treatment of an underlying infection associated with MZL (such as H. pylori for primary gastric ENMZL and C. jejuni for IPSID) is advisable in order to remove any inciting factor. Otherwise, as already mentioned, the same treatment approaches used for FL (see Chapter 82 ) are usually followed for ENMZL. Nonetheless, a few studies specifically addressing MZL are worth reviewing here.
Single Agent Rituximab
Given its low toxicity, single-agent immunotherapy with rituximab has generated a lot of interest in the management of advanced disease. Several small phase 2 studies have reported ORRs of around 75%, with up to 48% CR in patients without previous therapy. In one of the series, median time to treatment failure was 22 versus 12 months in chemotherapy-naive versus non-naive patients, respectively. Rituximab has activity in t(11;18)–positive disease. Maintenance rituximab is controversial, and no data are available to support its use in ENMZL. Previous studies failed to showed OS benefit of maintenance rituximab and accordingly it is not usually recommended.
Alkylating Agents
Single-agent alkylators have been used in this setting. A study of 24 patients with gastric ENMZL (stages IE or IV) treated with daily oral chlorambucil or cyclophosphamide for 12 to 24 months showed a CR rate of 75%, with the remaining patients achieving PR. Remissions were durable in approximately half of the patients after a median follow-up time of 45 months. Another study of 21 patients (stages I to IV) treated with continuous alkylating drugs documented a CR rate of 42% and 89% in t(11;18)–positive and t(11;18)–negative disease, respectively. After a median follow-up of 7.5 years, CR was sustained in all patients with translocation-negative disease, but in only one patient with t(11;18).
The combination of rituximab and chlorambucil has produced an impressive 100% CR rate in 13 patients with t(11;18)–positive gastric ENMZL (stages I to IV). No relapses were observed after a median follow-up of 24 months in this report but long-term data were not available. This combination has been studied in a randomized controlled trial versus each single agent in 393 patients not responding or not suitable for local therapy. CR rate, and 5-year event-free and PFS were better with the combination, but 5-year overall survival (OS) was 90% in the whole population, without significant differences between the three arms.
Responses to bendamustine (with or without rituximab) have been documented in patients with relapsed/refractory indolent lymphomas, a small fraction of which were ENMZL. A retrospective series of 14 ENMZL patients reported 10 CR and 3 PR with the combination of bendamustine and rituximab. MALT2008-01, a phase 2 study, reported excellent results in 60 patients treated with 4 cycles of bendamustine and rituximab, which resulted in 100% ORR and 98% CR/unconfirmed CR. The combination was active in patients with t(11;18) and most recently the results were sustained, with a median follow up of 7 years.
You're Reading a Preview
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here