Introduction

In 1983, Isaacson and Wright introduced the concept of mucosa-associated lymphoid tissue (MALT) lymphomas [1]. MALT lymphomas are currently recognized, according to the World Health Organization (WHO) classification, as a distinct category of low-grade B-cell lymphoma with unique pathogenic, cytogenetic, histological, and clinical features [2]. MALT lymphomas may arise at any anatomical site. The most common and best-studied localization is the stomach [3]. This paper reviews the main features of MALT lymphomas and their current treatment approaches.

Epidemiology

MALT-lymphoma is the third most common non-Hodgkin lymphoma in the Western world, accounting for approximately 8 % of all non-Hodgkin lymphomas [4]. In a recent Surveillance, Epidemiology, and End Results (SEERS) database analysis in a total of 15,908 patients with marginal zone lymphoma (MZL), MALT lymphoma was the most common subtype, with an incidence rate of 1.59 per 100,000 adults and constituting 5 % of all B-cell lymphomas [5]. The incidence increases with age and the median age of occurrence is around 60 years with a slight female predominance. MALT lymphomas may arise at any extranodal site. The stomach is the most frequently involved localization, representing more than one-third of MALT lymphoma cases [6]. Other frequently involved sites include: salivary glands, ocular adnexa, skin, lung, and small intestine. There is a geographic variability of the various anatomic locations, mainly reflecting different epidemiological risk factor distribution.

Epidemiological Associations of Various Factors and the Development of MALT Lymphomas

Infectious Agents

Helicobacter pylori (Hp) infection has been detected in up to 90 % of gastric MALT lymphomas and represents the strongest association between a causative agent and lymphomatogenesis [7].

Chlamydophyla psittaci (Cp) has been associated with the development of MALT lymphomas in ocular adnexa [8]. Its prevalence displays marked geographic variability. Studies from Italy, Austria, Germany, and Korea have reported a high prevalence of Cp in ocular adnexal lymphoma (OALs), ranging from 47 to 80 %, while studies from the United States have not reproduced these results [8].

Borrelia burdgorferi (Bb) has been associated with the development of cutaneous MALT lymphoma. A geographical variation has also been reported in the prevalence of Bb in cutaneous MALT lymphomas, ranging from 10 to 42 % in endemic areas such as Europe, while being absent in non-endemic areas [8].

Cambylobacter jejuni (Cj) has been implicated in the pathogenesis of immunoproliferative small intestine disease (IPSID), a variant of MALT lymphoma [9].

Autoimmune Disorders

There is a strong association between Sjogren syndrome (SS) or lymphoepithelial sieladenitis and the development of MALT lymphomas, mainly those affecting the salivary glands [10]. Patients with SS have a 44-fold increase in the risk of developing a MALT lymphoma.

Hashimoto’s Thyroiditis represents a common background for the development of thyroid MALT lymphoma, since in 94 % of thyroid lymphomas there is evidence of underlying lymphocytic thyroiditis [11]. Patients with Hashimoto’s thyroiditis have a three-fold excess risk of developing lymphoma.

Pathogenesis

MALT lymphomatogenesis is a multistep process initiated by chronic antigenic stimulation, which leads to the assembly of polyclonal B-lymphocytes in extra nodal sites and the formation of acquired MALT [12]. This chronic inflammation may be due to either chronic infection or the autoimmune processes [13]. Sustained antigenic or auto antigenic stimulation promotes not only polyclonal B-cell proliferation, but also triggers the assembly of various microenvironmental components to the site of inflammation [14••] and may promote the selection of auto-reactive B-cell clones. Genomic instability eventually may lead to the prevalence of a monoclonal B-cell population (Figure 1) [12, 15]. Gastric MALT lymphomas represent the best-studied model for MALT lymphomatogenesis. In the early 1990s, Wotherspoon et al. first described a high incidence of Hp gastritis among patients with gastric MALT lymphoma and that the acquisition of gastric MALT is developed in Hp-infected individuals, which in turn facilitates the development of gastric MALT lymphoma [7]. An inflammatory microenvironment plays an important role in the pathogenesis of these lymphomas, underscoring the notion that MALT lymphomas are highly dependent on external signals. The most important inflammatory components include T-lymphocytes, neutrophils, macrophages, and endothelium [14••, 16]. T-cells act via two major mechanisms. The first one includes the CD40-CD40 ligand axis and the expression of Th-2 type cytokines and costimulatory molecules such as CD86, while the other involves the recruitment of regulatory T-cells (T-regs) (CD4+, CD25+, FOXP3+) by tumor B-cells via secretion of the T-regs-attracting chemokines (CCL17, CCL22) [17, 18]. Tumor cell proliferation is strongly enhanced by the presence of intratumoral CD4+ T cells. The significant role of T-regs is shown by the fact that the depletion of these cells can block tumor growth, at least in vitro [18]. Another important component of the inflammatory microenvironment is represented by neutrophils. As a part of the immune response to Hp infection, neutrophils produce reactive oxygen species (ROS), which in turn cause DNA-damage enhancing lymphomatogenesis [19, 20]. In particular, Hp strains positive to cytotoxin-associated gene A (CagA) generate a strong inflammatory response and ROS secretion [21]. CagA is translocated into human B-lymphocytes and acts as an oncoprotein that contributes to lymphomatogenesis by regulating intracellular signaling pathways [22]. Monocytes represent a reservoir for pathogens and support MALT lymphoma progression mainly by releasing a proliferating inducing ligand (APRIL) upon Hp infection [23]. Finally, in the endothelia, several chemokine and chemokine receptors are very important compounds in the inflammatory milieu [24–29]. MicroRNAs (miR) also may play a critical role in MALT lymphomatogenesis. MiRs are implicated in posttranscriptional gene regulation [30]. Various miRs have been shown to be either up- or downregulated in MALT lymphomas [31, 32]. Furthermore, hypermethylation of certain genes such as p16, MGMT, and MINT31 may be an important epigenetic mechanism in the development of MALT lymphomas [33–35]. In conclusion, the development and progression of early-stage MALT lymphoma is based on the cooperation of B-cell receptor-derived signals and T-helper cell signals in cooperation with other components of the inflammatory microenvironment in some genetically predisposed patients.

Fig. 1
figure 1

Main pathogenetic pathways of gastric MALT lymphoma

Full size image

Immunogenetics

The lymphoma cells express surface rearranged, somatically mutated immunoglobulins in the vast majority of MALT lymphomas [36–39]. However, in rare cases, mainly those derived from primary lung localization, the IGHV genes are somatically unmutated [40]. Approximately half of the cases show evidence of intraclonal variation and positive and/or negative selection [36–38, 41]. BCR immunoglobulins show a pattern of polyreactivity, displaying equally strong affinity toward a diverse panel of foreign, but also self-antigens [42]. Moreover, there is an apparently biased usage of certain IGHV gene segments by MALT lymphomas originating in specific anatomical sites [36, 43–47].

Chromosomal Abnormalities

Although antigen stimulation is important for early lymphomatogenesis, it becomes less relevant as the disease progresses, probably due to the acquisition of additional genetic aberrations [48]. Chromosomal rearrangements lead to constitutive activation of the nuclear factor –κΒ (NF-κB) signaling pathway, thus providing tumor B-cells with BCR-independent proliferation [48–51]. The most common recurrent translocations are presented by t (11; 18), t (14; 18), t (1; 14) and t (3; 14). Their frequency, biologic role, and anatomical distribution are presented on Table 1. The t (11;18) is the most common translocation found in MALT lymphomas, involving up to 25 % of gastric MALT lymphoma and 40–60 % of MALT lymphomas occurring in the small intestine and pulmonary location [21, 49]. It results in the production of a chimeric oncoprotein API2-MALT1, which leads to the constitutive activation of NF-Κβ, providing tumor cells with antigen-independent activation and survival. This is in agreement with the commonly seen resistance of MALT lymphomas carrying this translocation to antibiotic therapy [52]. The translocation is often associated with disseminated disease and submucosal involvement. It is more frequent in Hp-negative cases, is absent in Hp-positive gastritis, but is often found in gastric MALT lymphoma patients infected with CagA-positive Hp strains [21]. Despite the initial notion that it is infrequently associated with diffuse, large B-cell lymphoma transformation, recent data suggests that t (11; 18) can be found at approximately equivalent frequencies in both gastric MALT lymphoma and gastric DLBCL [53]. t (11; 18) is highly specific for MALT lymphoma and it is rarely seen with other cytogenetic abnormalities [54].

Table 1 The most common recurrent chromosomal abnormalities of MALT lymphomas
Full size table

The different pattern of cytogenetic abnormalities in the various anatomic locations probably implies distinct processes of lymphomatogenesis [55].

Besides the aforementioned recurrent translocations, several other genetic aberrations have been described in MALT lymphomas with varying frequency. Numerical chromosome abnormalities, mainly trisomies, are commonly detected in MALT lymphomas, but are nonspecific. Trisomy 3q27 is the most common chromosomal abnormality in gastrointestinal (GI) lymphomas, being present in about 50–60 % of cases [56]. Trisomy 18 may be seen either alone or in association with trisomy 3, and both of them are associated with advanced disease stage. The pathogenic role of these two gains is still undefined. The 6q23.3 deletion is another frequently recurrent genetic abnormality, which occurs in approximately 15–30 % of cases [56]. It affects the TNFAIP3 (tumor necrosis factor, alpha-induced protein 3) gene, also known as A20, which is involved in limiting the activity of the NF-Κβ pathway. Inactivation of A20 also occurs by mutations, similarly to SMZL. In contrast to chromosomal translocations, unbalanced genomic aberrations display no preferential anatomical distribution. Other less frequent genetic aberrations include: del17p, +8q, and +6p. Several other somatic genetic alterations have been described in the literature with varying frequencies, mainly depending on the anatomical localization as well as the geographical locales [57].

Diagnosis

Tissue Sampling and Macroscopic Findings

The diagnosis of MALT lymphomas is highly dependent on the sampling of a representative tissue. This is especially important in gastric MALT lymphomas since they are usually mimicking chronic gastritis.

Histopathological Findings

The establishment of diagnosis is based entirely on the histopathological and immunohistochemical examination. In early 1980s, Isaacson and Wright observed that certain lymphomas of the GI tract recapitulated features of the MALT of Peyer patches [58]. MALT lymphomas rarely arise in sites of native MALT.

Histopathology

Tumor cells infiltrate the marginal zone of reactive B-cell follicles and spread out to form larger confluent areas and sometimes colonize germinal centers of the reactive follicles, resulting in a nodular pattern. Lymphoma cells often invade and destroy the glandular epithelium with the formation of the lymphoepithelial lesions (LEL). These are defined as aggregates of more than three lymphoma cells within the glandular epithelium. Even though LEL are characteristic, they are not specific for MALT lymphomas, since they can also be present in reactive conditions as well as in other lymphomas. The lymphomatous population is heterogeneous, consisting of small lymphocytes, centrocyte-like cells, monocytoid, and plasmacytoid cells. In some cases plasmacytic differentiation is prominent. Large cells are scattered within the lymphoid population, usually comprising less than 20 % of the total population. In rare cases in which the blast cells form solid or sheet-like proliferations, the diagnosis should be that of a diffuse, large B-cell lymphoma. Besides the tumor cells, a reactive population is also recognized, consisting mainly of T-lymphocytes, neutrophils, and monocytoid cells [2].

Immunohistochemistry

Lymphoma cells typically express IgM, less often IgA or IgG, and are usually IgD-negative. Their immunophenotype recapitulates that of marginal zone cells. They are positive for CD20, CD79a, BCL-2, CD21, and CD35, while they are negative for CD5, CD10, CD23, and cyclinD1. Expression of CD11c is variable, while CD43 is expressed in approximately 30–50 % of cases. The demonstration of immunoglobulin light chain restriction helps to differentiate MALT lymphomas from a reactive lymphoid infiltrate [2]. Recently, two new immunohistochemical markers have been reported to be characteristic for MALT lymphomas: immunoglobulin receptor translocation associated-1 (IRTA-1) and myeloid cell differentiation nuclear antigen (MDNA). These molecules are also expressed in other MZL, although IRTA-1 is not expressed in SMZL [59, 60].

Detection of Monoclonality by PCR

In some instances it is difficult to differentiate MALT lymphomas from reactive conditions. In such cases, the detection of monoclonality by polymerase chain reaction (PCR) may help the differential diagnosis. PCR, however, cannot replace the histopathologic examination. Monoclonality may be detected in some cases of chronic inflammation, such as myoepithelial sieladenitis, while in up to 15 % of overt MALT lymphomas monoclonalty is not detected by PCR [61]. Consequently, PCR can be used complementary to pathological examination and never alone for the establishment of MALT lymphoma diagnosis.

Clinical Features

Patients usually are asymptomatic or present with symptoms related to the primary organ involvement, e.g., for gastric MALT lymphomas the most common presenting symptoms are dyspepsia, nausea, and epigastric pain. Usually there is no impairment of patient performance status. B-symptoms are extremely rare, and adverse biological prognostic factors such as high lactate dehydrogenase (LDH) or β2-microglobulin levels are infrequently elevated [62, 63]. Despite the initial notion that MALT lymphomas have a tendency to be localized at presentation, recent data suggest the opposite, since more than one-third of the cases display involvement of multiple extranodal sites [62–67]. There is a higher tendency for multiple MALT involvement in extragastric MALT lymphomas compared to gastric ones. Dissemination at presentation has been reported with varying frequencies, probably relating to the extensiveness of the staging procedures [64, 66]. Bone marrow is not usually infiltrated (2–15 %). Generalized lymphadenopathy is also rare (<10 %). The frequency of paraproteinemia varies depending mainly on the primary MALT localization.

Prognosis

MALT lymphomas usually run an indolent clinical course with long survival. In a retrospective study by the International Extranodal Lymphoma Study Group (IELSG) of a large series of patients with non-gastric MALT lymphoma, the five-year overall survival (OS) was 90 %, despite the fact that one-quarter of cases presented with advanced disease stage and regardless of treatment type [62]. A recent study evaluated the clinical characteristics and survival of patients with MZL subtypes in the unselected population included in the SEER database [52]. The median OS was 12.6 years with a significant improvement in the probability of lymphoma-related death for patients with MALT lymphoma that were diagnosed after 2000, while further improvement was observed after 2005. There is no prognostic index specific for MALT lymphomas. The International Prognostic Index (IPI) has been correlated with time to relapse in some studies, while the utility of Follicular Lymphoma IPI (FLIPI) is controversial [68]. The tendency of MALT lymphomas to involve multiple MALT sites at presentation is not associated with inferior outcome, as it has been shown by various studies [62–67]. On the contrary, bone marrow and nodal involvement adversely affected overall survival, at least in some series [62]. In contrast to extragastric, gastric MALT lymphomas are presented as localized disease in the majority of cases (~90 %) [63]. The median time to progression is apparently better for the gastric compared with the nongastric lymphomas (9 vs. 5 years, respectively), but no significant differences in OS have been shown [66]. Despite frequent relapses, MALT lymphomas most often maintain an indolent clinical course. Primary disease localization may have an impact on outcome [52, 62, 66]. Histologic transformation to large-cell lymphoma occurs less often than in follicular lymphomas (<10 %) as a late event [64, 67].

Staging Procedures

Gastric MALT Lymphomas

The staging procedures at diagnosis should include an esophagogastroduodenoscopy (EGD), with multiple biopsies not only from sites with abnormal appearance but also from every region of the stomach, duodenum, and gastroesophageal junction [69•]. In addition to routine histology, fresh biopsy should be available for cytogenetic studies, particularly for the detection of t (11; 18) translocation. The presence of Hp infection must be determined either by histology or serology studies. Endoscopic ultrasound is recommended to evaluate the depth of gastric wall infiltration and regional lymph node involvement. Other studies should include the typical procedures followed for staging any lymphoma (e.g., CBC, biochemical studies, whole-body CT scan, bone marrow biopsy). Bone marrow involvement has been reported in up to 15 % of cases. The value of a positron emission tomography is controversial and currently is not recommended in the routine work-up of MALT lymphomas patients [69•]. The best staging system is still not known. The “Paris staging” can adequately record the tumor extension and depth of infiltration [70].

Non-Gastric MALT Lymphomas

Since extragastric MALT lymphomas usually present with multiple MALT involvement, it is important to assess all MALT sites at presentation. Besides the typical staging procedures, further investigation should be guided by the clinical presentation. However, the extent of staging procedures is still controversial. Our strategy is to perform upper endoscopy in all extragastric MALT lymphoma cases in addition to the aforementioned investigation, since the stomach represented the most commonly seen secondary MALT site of involvement in our series [66]. Extensive staging is very important, especially when local treatment approaches are going to be used. Moreover, it is highly recommended to search for the detection of certain chronic infections that may have a pathogenic role. The Ann Arbor staging system has limited value in staging MALT lymphomas [62]. However, there is no alternative staging system for extragastric MALT lymphomas.

Treatment

Gastric MALT Lymphomas

First-line therapy for gastric MALT lymphoma associated with Hp infection is well documented, and includes the eradication of Hp with antibiotics and a proton pump inhibitor (PPI) [7]. Several effective antihelicobacter regimens are available [71]. The most commonly used regimen includes a combination of PPI plus clarithromycin-based therapy with either amoxicillin or metronidazole for 10-14 days. In cases with persistent Hp infection a second-line treatment should be used with alternative triple- or quadruble-therapy regimens of PPI plus antibiotics. Successful eradication of Hp is associated with histologic regression of the gastric MALT lymphoma in approximately 75 % of patients with localized disease [72, 73••].

Main factors predicting adverse response to eradication include:

  1. 1.

    Absence of Hp infection, although eradication therapy may also be successful in a small proportion of Hp-negative patients (~15 %) [74].

  2. 2.

    Infiltration beyond the mucosa layers and regional lymph node involvement. Complete endoscopic and histopathologic remission is higher in stage I than in stage II disease, and when lymphoma is confined to the submucosa, as compared to a deeper invasion [73••].

  3. 3.

    Translocations t (11; 18) and t (1;14) are associated with resistance to Hp eradication therapy, although complete lymphoma regression can still be obtained in about 20 % of t (11; 18)-positive cases after Hp eradication [73••].

  4. 4.

    Proximal location has been associated with inferior response rather than localization in the distal stomach [73••].

  5. 5.

    Patient ethnicity. Asian patients respond better than Western patients [74].

In addition, increased numbers of Tregs in gastric lymphoma may be associated with a better response to Hp eradication therapy [75•]. Furthermore, the over-expression of either miR-142-5p or miR-155 in MALT lymphoma was associated with a lower probability of response to Hp therapy [76].

According to the currently published ESMO guidelines, Hp eradication therapy should be given to all gastric MALT lymphomas, independently of stage and even in Hp-negative cases [69•].

Follow-up of Patients After Eradication Therapy

Lymphoma may take more than 12 months to regress after Hp eradication, indicating that refractoriness should not be assumed prematurely, even in cases with residual lymphoma at the histological level [69•]. Eradication of Hp should be checked for at least six weeks after the end of therapy. Lymphoma regression should be checked by EGD by 3–6 months after eradication therapy. Patients who have achieved at least a clinical and endoscopic remission should be followed closely with EGD every 3–6 months, since disease may regress or remain stable in the majority of patients. Detection of monoclonality by PCR has no clinical relevance, since it has been shown by several studies that the persistence of monoclonal B cells is found in almost half of patients with histologic regression of MALT lymphoma [61, 77]. The interpretation of post-treatment gastric biopsies may be difficult since there are no uniform reproducible criteria for the definition of histologic remission. The recently proposed GELA system may help to solve this problem, although its reproducibility requires further confirmation [78]. Patients in complete remission require regular long-term follow-up endoscopies, not only due to the risk of lymphoma recurrence, but also due to the risk of development of gastric adenocarcinoma [79]. Endoscopic check-up should be performed every six months for the first two years and every 12–18 months thereafter.

Outcome

Complete remissions achieved after Hp eradication therapy are usually prolonged. Relapses are in the range of 5–17 % in reported series, may occur either with or without Hp infection recurrence, and essentially involve the stomach [73••]. In an analysis of 994 patients, lymphoma relapse occurred in 7.2 % after 3,253 patient-years of follow up, with a yearly recurrence rate of 2.2 % [73••]. Several trials showed that the five-year OS and disease-free survival (DFS) rates in patients with early disease stage were 90 % and 75 %, respectively [80]. The risk of transformation into aggressive lymphoma is lower than in other indolent lymphomas, although it has not yet been clearly determined.

Treatment of Patients With Gastric MALT Lymphoma who Fail Antibiotics Treatment as Well as Those With non-Gastric MALT

There is no standard treatment approach for the management of these patients. Primary MALT localization is an important factor due to organ-specific problems requiring particular treatment strategies.

Local Therapies

Surgery is an option in some cases, such as in thyroid MALT lymphoma, with excellent long-term, progression-free survival. For gastric MALT lymphoma systemic therapies or radiotherapy is suggested as first-line treatment, while gastrectomy is not superior to organ preservation strategies and is no longer considered as a first-line treatment option [69•].

Radiotherapy with moderate doses of irradiation (24–30 Gy) has extensively been used and is associated with high response rates (>90 %) and long response duration [81, 82]. Intralesional rituximab in primary cutaneous MZL proved to be highly effective, leading to favorable outcome of the disease without the need of systemic treatment or local radiotherapy [83].

Systemic Therapies

Systemic therapies including monotherapy or combination chemotherapy and immunotherapy, are effective in patients with gastric or non-gastric MALT lymphomas. There is no definitive data to support the choice of one chemotherapy regimen over the other. Single-agent alkylators, mainly chlorambucil, have shown clinical activity with acceptable toxicity, although t (11;18) has been associated with refractoriness to alkylating agents [84]. Nucleoside analogues, mainly cladribine, have been used in phase II studies demonstrating significant activity with more than 80 % complete responses, albeit with significant toxicity [85]. Combination chemotherapy with chlorambucil plus mitoxantrone and prednisone, as well as fludarabine in combination with mitoxantrone and the classic CVP are active and well-tolerated regimens [86, 87]. Rituximab monotherapy is also effective in gastric MALT lymphomas regardless of t(11;18) with an overall response rate of 77 %, however, complete responses have been observed in about only half of the cases [88]. Levy et al. showed a significantly greater response to rituximab-chlorambucil than to rituximab alone in patients carrying the t(11;18) [89]. A recent randomized study by the IELSG found that the addition of rituximab to chlorambucil was significantly better than chlorambucil alone in terms of event-free survival, with no OS benefit [90•]. Aggressive anthracycline-containing regimens are to be reserved for cases with transformation or bulky masses [91]. Antibiotic treatment with doxycycline appears to be a reasonable first-line therapy for patients with OAL [92]. Antibiotics, however, remain experimental for the time being in patients with other non-GI MALT lymphomas [93].

Overall, the results of these studies suggest that both radiotherapy as well as systemic therapies are effective. For practical reasons, we suggest a treatment algorithm of MALT lymphomas based on the available data (Figure 2).

Fig. 2
figure 2

Suggested MALT lymphoma treatment algorithm

Full size image

Novel Agents

Bendamustine alone or in combination with rituximab has shown significant activity in indolent lymphomas, including MALT lymphomas [94]. A phase II study demonstrated significant antitumor activity of lenalidomide [95]. Furthermore, the combination of lenalidomide with rituximab resulted in a 90 % response rate [96]. Besides rituximab, other anti-CD20 moAbs such as ofatumumab and obinutuzumab (GA101) have been recently tested in several clinical trials in patients with relapsed/refractory indolent lymphomas with promising results [97, 98]. Novel therapies such as BTK-inhibitors, vorinostat, and PI3-K inhibitors are in clinical trials for patients with relapsed or refractory MZL [99, 100]. Everolimus displayed moderate activity in a phase II clinical trial including patients with relapsed/refractory MZL [101]. Bortezomib was also associated with moderate activity and significant toxicity in a phase II study [102].

Conclusions

Extraordinary progress has been made regarding the understanding of pathogenesis and clinical behavior of MALT lymphomas. Chronic inflammation is usually the necessary background for the development of a MALT lymphoma, while the tissue microenvironment has a significant role in the establishment of the neoplastic process. Patients with MALT lymphoma have a favorable outcome without significant difference between GI or non-GI lymphoma or between localized and disseminated disease. With the exception of Hp-positive gastric MALT lymphoma, there are no treatment guidelines for the management of patients with MALT lymphoma. Therefore, controlled clinical trials are needed in order this goal to be achieved. International study groups are significantly contributing to this effort.