Insights, Knowledge Gaps, and Priorities in Marginal Zone Lymphoma Research

Abstract

Marginal zone lymphoma (MZL) is a rare, indolent form of non-Hodgkin lymphoma that arises from B cells in the marginal zone of lymphoid tissues. MZL comprises 3 key subtypes: extranodal, nodal, and splenic MZL. Despite being generally slow growing, MZL presents significant challenges due to its heterogeneous nature, inconsistently defined disease, and the limited efficacy and availability of current treatments. Advancements in targeted therapies and a deeper understanding of the molecular underpinnings of MZL are critical to improving patient outcomes and achieving more durable remissions. At the Lymphoma Research Foundation’s 2024 Marginal Zone Lymphoma Virtual Scientific Workshop, researchers gathered to discuss recent developments in both basic scientific and clinical research so that together we can continue to develop our understanding of MZL and improve outcomes for patients. This report, which includes a summary of each presentation, aims to review the findings presented at the workshop. Additionally, it highlights opportunities, reviews questions, and assesses areas for future study to set the stage for treatment advancements in the coming decades.

Cochairs of the MZL Workshop

Introduction

Marginal zone lymphoma (MZL) is an indolent B-cell non-Hodgkin lymphoma that originates in the marginal zones of lymphoid tissues, encompassing the key subtypes of extranodal, nodal, and splenic MZL. Accurate diagnosis and effective prognostication are challenging due to MZL’s heterogeneous presentation and overlapping features with other lymphomas. Current treatment options are limited and often yield only partial remissions, highlighting the need for more effective and targeted therapies. Improved diagnosis and disease characterization are essential for optimizing outcomes, and recent research has made significant strides in understanding the genetic and molecular landscape of MZL. These advancements are poised to improve patient outcomes by enabling more precise diagnostics, prognostics, and therapeutic interventions.

Recognizing the need for accelerated MZL research, the Lymphoma Research Foundation has provided MZL-specific research grants and developed an MZL steering committee, a working group that includes both basic scientists and translational/clinical researchers from North America and Europe. Since April 2019, the group has met regularly to allow researchers to share their work and offer a unique opportunity for collaboration among investigators across a wide range of MZL areas of interest. Through this type of exchange, thoughts on the current and future direction of MZL research are shared, and researchers are provided with a unique opportunity to develop collaborations needed to continue to drive MZL research forward.

The 2024 MZL Virtual Scientific Workshop, held on May 3 and 4, 2024, included sessions on MZL pathology; molecular taxonomy; viral, microbial, and antigen factors linked to MZL; developmental therapeutics; MZL epidemiology, prognosis, and transformation; criteria for assessment and evaluation of response; an international overview of MZL clinical trials; and an open forum to establish a road map for MZL research priorities in the short (1-5 years) and long (5 years or more) terms.

The State of MZL Since 2019

To kick off the workshop, Davide Rossi, MD, PhD, deputy head of the Division of Hematology of the Oncology Institute of Southern Switzerland and head of the Laboratory of Experimental Hematology at the Institute of Oncology Research in Bellinzona, Switzerland, provided an overview of the research advancements that have occurred in MZL since the 2019 MZL workshop. Rossi utilized data from a PubMed search to illustrate that the annual number of MZL publications hovers around 400. Among the new publications, Rossi identified 86 studies with transformative research, most of which (n = 62) were clinical studies that fell into the following categories: staging and restaging, treatment, prognosis, and resistance. A total of 24 studies were translational and covered topics including predisposition, classification, genetics, microenvironment, and transformation. Given this background of new literature, Rossi expressed great excitement for the future of MZL research.

Next, Thomas Habermann, MD, professor of medicine at Mayo Clinic, Rochester, Minnesota, provided a road map and overview of MZL. Habermann revisited concerns and unanswered questions generated in the 2019 workshop and the corresponding long- and short-term solutions proposed to address those needs. The questions and concerns were categorized into the following sessions: biology and pathology; epidemiology and transformation; assessment criteria, response evaluation, and surrogate end points in MZL; MZL targeted pathways; etiology and natural history of MZL subtypes; and treatment of MZL. The following sessions share the community’s progress toward understanding more about each of these important categories.

Session I: Pathology

To open this session, Andrew Wotherspoon, MB BCh, FRCPath, consultant histopathologist at the Royal Marsden Hospital, discussed the gray areas of diagnosing MZL (Figure 1). Currently, diagnosis of MZL is based on analysis of peripheral blood and bone marrow aspirate/biopsy in combination with evaluation of molecular disease characteristics, but some difficulties with differential diagnoses persist. Wotherspoon covered 6 specific challenges in MZL diagnosis, the first of which was differentiating between splenic MZL and other primary splenic small B-cell lymphomas. Wotherspoon’s approach for differentiating between hairy cell leukemia, splenic MZL, and splenic diffuse red pulp small B-cell lymphoma relies on assessment of peripheral blood appearances, the degree of intrasinusoidal disease, and differentiating molecular findings. To distinguish between primary and secondary MZL in lymph nodes, Wotherspoon suggests searching for extranodal primary MZL only if a patient has stage I nodal MZL. Differential diagnosis of early MZL vs reactive lymphoid tissue can be difficult; useful differentiators that are often (but not always) indicative of early MZL overreactive tissue include lymphoepithelial lesions that are well formed and have eosinophilic appearance, the presence of Dutcher bodies, centrocyte-like cell morphology and cytological atypia, expression of CD43 and CD5, and light chain restriction. Clonality assessments may be helpful but should be interpreted with caution. Indicators of lymphoplasmacytic lymphoma/Waldenström macroglobulinemia (WM) vs splenic MZL include levels of immunoglobulin M (IgM), marrow infiltration pattern, the presence of mast cells, plasma cell count, and some immunohistochemical factors. Differentiating between extranodal MZL and follicular lymphoma (FL) can be aided by assessing the infiltrate pattern, determining the presence of lymphoepithelial lesions, and careful immunophenotyping. Atypical marginal zone hyperplasia exhibits distorted follicular structures with infiltration in the marginal zone area. Features differentiating hyperplasia from MZL include aberrant CD43 expression, CD27 negativity, light chain restriction, λ light chain restriction, high proliferation, location (tonsil and appendix are most common), age, and clonality. The insights provided in this talk may help physicians and pathologists discriminate between difficult MZL diagnoses.

FIGURE 1. Key Molecular Alterations in NMZL and SMZL With NNK and DMT Genotypes

This is a schematic representation of the genes and pathways that are molecularly deregulated. The prevalence of these alterations is indicated alongside each gene or pathway, providing an overview of frequency and relevance.

Next, James Cook, MD, PhD, professor of pathology at the Cleveland Clinic, discussed the existence of MYD88-negative WM and MYD88-positive MZL (Figure 2). In general, distinguishing between MZL and lymphoplasmacytic lymphoma (LPL) is challenging due to their overlapping clinical, morphological, and immunophenotype features. A 2012 study identified MYD88 mutations in the vast majority of bone marrow WM samples and many LPL samples, leading to the recognition of the MYD88 L265P mutation as a common but not exclusive feature of LPL (95%-97% of cases).¹ A small percentage of patients with LPL are reported to have wild-type MYD88; MYD88 wild-type LPL cells appear to have similar characteristics to MYD88-mutated LPL cells. MYD88 mutations are not exclusive to LPL and can also occur in other lymphomas. In MZL, up to 10% of patients have been reported to harbor mutations in MYD88; however, most data on MYD88 mutations in MZL come from patients with splenic disease and bone marrow biopsies, and most reported cases have plasmacytic differentiation and IgM protein, which makes the true disease for these samples difficult to ascertain. The MYD88 L265P mutation has been reported in a small percentage of nodal and splenic MZL cases, and in mucosa-associated lymphoid tissue (MALT) lymphomas, MYD88 mutations have been reported in gastric and ocular adnexa sites. Overall, Cook concluded that MYD88-negative WM and MYD88-positive MZL cases do seem to exist but are rare. In LPL, additional MZL biomarkers would be helpful to better understand MYD88’s relationship to MZL. In splenic MZL, the MYD88 L265P mutation appears rare to absent; in extranodal MZL, MYD88 L265P does occur at certain sites; and while there are nodal MZL cases with MYD88 L265P, the criteria for nodal MZL need refinement. The presence of MYD88 L265P favors a diagnosis of LPL over MZL but is not 100% sensitive or specific. While the diagnosis of extranodal MALT lymphoma is generally straightforward, distinguishing between splenic and nodal MZL from LPL remains a challenge.

FIGURE 2. Anatomical Representation of Various MZL Subtypes.

This figure illustrates the anatomical distribution, molecular characteristics, and associated triggers of various subtypes of MZL.

Session II: Molecular Taxonomy

Ming-Qing Du, PhD, MB, FRCPath, professor of oncological pathology, Division of Cellular and Molecular Pathology at the University of Cambridge, discussed the genetic and immune characteristics of extranodal MZL. Du reviewed key genetic changes that have been implicated in MALT lymphomas. It has been established that marginal zone B-cell differentiation is largely driven by transcription factor signaling, but genetic disease varies by disease site. For example, gastric MALT lymphoma is characterized by Helicobacter pylori–specific T-cell signaling. In thyroid MALT lymphoma, genetic changes affect B-cell and T-cell function via inactivation of PD-L1 and TNFRSF14 (herpesvirus entry mediator) signaling and the inactivation of FAS ligand signaling. In salivary MALT lymphomas, notable genetic changes include mutations (in the G-protein coupled receptor GPR34 and chemokine receptor CCR6) that promote lymphatic transcription programs. CCR6 signaling has also been implicated in gastric MALT; however, Du noted that CCR6 signaling is independent of genetic changes to the receptor and is commonly maintained by ligand stimulation in inflammatory conditions (ie, H pylori infection). Du summarized the discussion by showing how similar mechanisms drive malignancy in the different MALT disease sites but involve different players: The H pylori–specific T cells in gastric extranodal MZL, the exaggerated T-cell function in thyroid extranodal MZL, and the enhanced GPCR signaling in salivary extranodal MZL all lead to increased transcription of MZL-related genes.

Next, Anne J. Novak, PhD, consultant, Division of Hematology, Department of Internal Medicine; consultant, Department of Immunology; and professor of medicine at Mayo Clinic, described the genomic, transcriptomic, and biologic characterization of MZL. Novak discussed efforts to use next-generation sequencing strategies to identify shared biology and disease mechanisms across B-cell lymphoma subtypes. Using acquired tumor samples from 64 B-cell lymphomas, Novak’s group performed bulk RNA sequencing, tumor-normal whole exome sequencing, and immune profiling to identify distinct clusters of patients with differences in event-free survival (EFS) and overall survival (OS).² The 5 patient clusters had distinct biological, genetic, and immune features. A gene expression signature including 113 genes was identified and associated with inferior EFS and OS in low-grade B-cell lymphomas; using patient data, the gene signature identified those with significantly worse EFS and OS. Using the mining algorithm for genetic controllers, a tool for predicting transcription factors of gene sets, DEK was identified as a potential regulator of the genes included in the predictive gene signature. DEK expression was associated with aggressive disease in low-grade B-cell lymphoma and correlated with cell cycle gene expression. Cells collected from more aggressive tumors showed higher levels of DEK expression, and in cell-based experiments, DEK depletion inhibited proliferation and was accompanied by reduced expression of cell cycle genes, reduced Bcl-2 and Bcl-xL expression, and increased p53 expression. DEK knockout cells also showed increased susceptibility to apoptotic agents. Future research will continue to explore the role of DEK in lymphoma.

Alexandar Tzankov, MD, surgical pathologist and head of the Department of Histopathology and Autopsy at the Institute of Medical Genetics and Pathology at University Hospital Basel, University of Basel, followed with a discussion of nodal MZL from a pathologist’s point of view. Nodal MZL is a diagnosis made in the absence of indicators of extranodal or splenic disease. Increasing age is the most important risk factor for nodal MZL. Histopathology often shows a nodular, inside-out pattern from the germinal center, occasionally exhibiting a blastoid morphology, and in some cases, the lymphoma cells have plasmacytoid differentiation. The nodal MZL phenotype is not very specific; samples may be positive for a range of immunohistochemical markers; the most common are Bcl-2, CD19, CD20, CD22, and CD79a. Tzankov noted that bone marrow involvement of nodal MZL is likely underestimated. There are no specific cytogenetic signatures of nodal MZL, and molecular genetics generally overlap with other MZL subtypes, though PTPRD and BRAF are notable in nodal disease, and mutations in MYD88 are rare. Genes mutated in nodal MZL are generally involved in chromatin remodeling, NOTCH signaling, and the p53 pathway. Transformation is poorly defined in MZL; it is often evidenced by the appearance of sheets of blasts. Other indicators of transformation may include the Ki-67 score, karyotype, and mutations in NOTCH3, TP53, andTBL1XR1 (discussed further by Luca Arcaini on day 2).To summarize, Tzankov explained that in practice, the diagnosis of nodal MZL requires a thorough analysis of lymphadenectomy samples, clinical and radiologic data, and bone marrow biopsies. Mutations in several genes related to MZL can provide diagnostic and prognostic information when used in the context of other disease information, and genetic aberrations provide opportunities to identify novel drug targets for MZL treatment.

Session III: Searching for New Pathogens/Antigens Associated With MZL

To open this session, Andrea Alimonti, MD, director of the Institute of Oncology Research, head of the Molecular Oncology Research Group, and full professor at Università della Svizzera italiana and ETH Zurich, discussed the microbiome of patients with lymphoma. In healthy individuals, the gut microbiome contains over 2000 different species of organisms with roles in metabolism, vitamin production, and xenobiotic and drug detoxification processes. Dysbiosis in the balance of the gut microbial environment can contribute to the onset of many diseases and is one of the canonical hallmarks of cancer. Microorganisms can induce tumorigenesis through a variety of mechanisms, including by inducing DNA damage, creating an inflammatory environment, or through secondary metabolite or hormone signaling. Gut bacteria can directly influence the onset and progression of cancer; the intratumoral microbiome is an ongoing subject of research in a number of different cancer types and has been implicated in metastatic colonization.

In non-Hodgkin lymphomas, specific bacteria have been linked to MALT lymphoma onset, including H pylori, Chlamydophila psittaci, Campylobacter jejuni, and Borrelia burgdorferi (Figure 3). In lymphoma, pathogen infections are proposed to promote malignancy via chronic stimulation of B-cell and T-cell lymphocytes and the formation of follicular lymphoid tissue, which subsequently leads to B-cell lymphoma. Antibiotic therapy has proven beneficial, especially in H pylori gastric MALT. The microbiota-gut-lymphoma axis presents exciting opportunities for new MZL prevention and treatment strategies via the modulation of gut microorganisms. Alimonti concluded with a summary of the wide array of therapeutic interventions for gut microbiome modulation, including dietary and supplemental interventions, fecal microbiota transplantation, engineered bacterial therapies, and phage therapy. A growing number of clinical trials are underway to explore gut microbial modulation to combat cancer.

FIGURE 3. Overview of the Pathogenetic Evolution of MZL.

This depicts the progression from polyclonal B cells to oligomonoclonal expansion, leading to antigen-dependent MZL and eventually antigen-independent MZL. The role of H pylori infection in chronic gastritis is illustrated, emphasizing the interaction between H pylori–specific T helper cells, B cells, and the activation of pathways such as NF-B via antigen stimulation (eg, CD40/CD40L interactions and cytokines such as BAFF). The figure also notes genetic alterations (eg, translocations involving MALT1, IGH, and BCL10, as well as TNFAIP3 inactivation) that contribute to constitutive NF-B signaling, underscoring the transition to antigen-independent lymphoma development

Following the microbiome discussion, Ethel Cesarman, MD, PhD, assistant director of the Molecular Hematopathology Laboratory of the New York-Presbyterian Hospital/Weil Cornell Medical College, shared an update on the interplay between viral infections and lymphoma. Viruses have a significant impact on cancers; an estimated 13% of cancers are considered to be caused by viruses.³ In lymphoma, the intersection between cancer, immunodeficiency, and herpesviral infection is of particular relevance. The Kaposi sarcoma–associated herpesvirus (KSHV, also known as human herpesvirus 8 [HHV8]) is associated with lymphoproliferative disorders including primary effusion lymphomas, extracavitary primary effusion lymphoma (PEL), and KSHV-associateddiffuse large B-cell lymphoma (DLBCL) not otherwise specified, and is linked to reactive lymphoid proliferation. PEL can occur in the spleen and other organs and is characterized by large tumor cells and characteristics of latency-associated nuclear antigen staining pattern. In multicentric Castleman disease, KSHV is always found in B cells that are expressing λlight chains; further exploration of this phenomenon may provide insight into pathogenic mechanisms. In tumor cells, KSHV expresses approximately 7 different genes involved in the virus latency program. Existing antiviral treatments are targeted at the viral lytic program, but the latency protein vFLIP has emerged as an attractive target for the treatment of KSHV-associated diseases due to its roles in autophagy and apoptosis.

Cesarman’s research group is working to identify small-molecule inhibitors of vFLIP for use in cancer treatment. Other viruses also express latent-phase proteins that play a role in carcinogenesis; latent-phase proteins expressed by Epstein-Barr virus (EBV) are involved in oncogenic processes in lymphoma. Pharmacologic screens have identified DNA methylation as a key regulator in this process, and preclinical experiments aimed at inducing latency followed by T-cell killing are showing promise as an antilymphoma treatment. Future studies will further investigate the potential for targeting the latency switch for treating EBV-positive cancers. Humans also carry endogenous viruses in their genomes that may play a role in oncogenesis. In an effort to characterize the endogenous retrovirome in lymphoma, Cesarman’s team is sequencing EBV-positive lymphomas and exploring their gene expression profiles to gain insight into their roles in the disease process. Viruses, both exogenous and endogenous, play an important role in lymphoma biology and continue to present exciting new avenues for drug development.

To wrap up the session, Nicholas Chiorazzi, MD, professor at the Institute of Molecular Medicine, Feinstein Institutes for Medical Research; Kanti R. Rai, MD, professor of molecular medicine at the Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, and professor of medicine at the Feinstein Institutes for Medical Research, highlighted the role of autoimmune B cells on chronic lymphocytic leukemia (CLL) disease biology and how these mechanisms might be investigated in MZL. Chiorazzi provided an overview of the research leading to the understanding that CLL is a disease of autoreactive B cells and that signaling through the B-cell receptor (BCR) is important to the development and evolution of the disease. In experiments to characterize the reactivity of the CLL B cells, bacterial strains were able to trigger reactivity with patient-derived CLL antibodies. Biochemical experiments suggest that a nonprotein bacterial antigen is responsible for this reactivity. Experiments with commercial antigen arrays and phage-display experiments have identified several antigens that are able to react with CLL antibodies and may be useful in guiding the development of new treatments. To perform these types of experiments in MZL, Chiorazzi suggests selecting BCRs from patients across the 3 major disease categories, expressing them as the patient’s isotype, and screening for natural antigens while considering the cell of origin and with the maturational pathways that marginal zone B cells follow. Tissue and protein arrays are available for these types of experiments and may lead to novel insights into MZL biology.

Session IV: Developmental Therapeutics

Alberto J. Arribas, PhD, of the Institute of Oncology Research, spoke about the deregulated pathways and potential vulnerabilities that can be exploited for drug development in MZL. Across the 3 MZL subtypes, several molecular pathways are commonly regulated, including chromatin remodeling, NOTCH, B-cell receptor, and NF-kB signaling. Outside of these commonalities, each subtype has its distinct profile of signaling pathway dysregulation. Arribas provided an overview of different promising therapeutic approaches targeting these various pathways in lymphoma, including Bruton tyrosine kinase inhibitors (BTKis) with and without anti-IL16 agents (under investigation in MZL and BTKi-resistant MZL); PI3K with and without STAT inhibitors, epigenetic drugs, and miRNA mimics (in MZL and PI3K inhibitor [PI3Ki]-resistant MZL); chimeric antigen receptor (CAR) T-cell therapy (in BTKi- and PI3Ki-resistant MZL); demethylating agents (in splenic MZL); bispecific T-cell engagers (in aggressive lymphoma); antibody-drug conjugates (in DLBCL and indolent lymphoma); immune checkpoint modulators including antibodies to CD47 (in refractory MZL and indolent lymphomas); IRAK4 inhibition with BTKi and PI3Ki (in MYD88-mutated disease); mTOR inhibition (in BTKi-, PI3Ki-, and Bcl-2 inhibitor–resistant MZL); Bcl-2 inhibition (in lymphoma); and CXCR4 inhibition (in MZL). Arribas concluded that BTK and PI3K remain interesting targets in MZL, though there is a need for approaches to overcome resistance, that multiple novel therapies are on the horizon, and that novel targeted therapies have potential applications in relapsed and refractory disease.

José Ángel Martínez Climent, MD, PhD, principal investigator of Lymphomas Group, Hemato-Oncology Program, Universidad de Navarra, presented the development and research applications of mouse models of MZL. The importance of the MYD88 mutation to B-cell lymphoma pathogenesis is well established, but Martinez-Climent noted that it is not known how the MYD88 mutation drives lymphomas with different clinical, histopathological, and immunological features. To investigate this question, Martinez-Climent’s research group crossed a murine line carrying MYD88^L252Pwith mice with genetic lesions in BCL2, BCR, P53, and BLIMP1. Comparison of the mice resulting from these crosses with data from patients with LPL and MALT lymphomas showed similarities in immunoglobulin secretion, indicating the mouse models recapitulate relevant disease characteristics. These models have provided useful information about MZL disease processes. In one case, Martinez-Climent’s research team had noticed that B cells and T cells colocalize with CD40 lymphocytes in biopsy samples, so to gain further insight, they performed single-cell RNA sequencing using murine-derived cells. The resulting data indicated that along with an increase in the number of B lymphocytes, T-cell accumulation increased with disease progression and sustained clonal B-cell survival. Functional assays using murine-derived tumor cells confirmed the importance of the B-cell and T-cell interactions for lymphoma cell survival. Blocking CD40 signaling decreased the viability of B cells in vitro and in mouse models, indicating the potential for disrupting this interaction as a therapeutic strategy. Other mouse models of MYD88/CD79B-mutated-DLBCL have also provided useful molecular information about the tumor microenvironment. Martinez-
Climent concluded that these mouse models serve as a proof of concept for advancing precision immunotherapy in B-cell lymphomas according to genetic and immunological characteristics.

Anastasios Stathis, MD, director of the Phase I Program, Oncology Institute of Southern Switzerland, and faculty member of Biomedical Sciences, Università della Svizzera italiana, closed the session with a review of ongoing phase 1 studies. From 2017 to 2023, 15 drugs have been approved by the FDA for the treatment of non-Hodgkin lymphoma, but the trials for these drugs enrolled low numbers of patients with MZL, if they enrolled any at all. Data from the Cancer Therapy Evaluation Program at the National Cancer Institute indicate that results from phase 1 trials show an overall rate of grade 5 adverse events of 1.81% and an overall response rate of 25.1%, including 43.2% in lymphoma.⁴ A systematic review of lymphoma also found an overall response rate higher than 30% in the majority of studies.⁵ Drug approvals in MZL have been limited and based on data from small trials. FDA-approved therapies for MZL include zanubrutinib (overall response rate, 68.2% [extranodal MZL, 64%; nodal, 76%; splenic, 66.7%]; complete response (CR), 25.8%) andlenalidomide (progression-free survival [PFS] not significant in the MZL cohort).^6,7 A total of 26 phase I trials are ongoing in lymphoma, though none are specific to MZL, including 16 phase 1 and 10 phase 1/2 trials. Twenty of these trials are assessing monotherapies, and 6 are evaluating combination treatments. Among the trials assessing small molecules are those targeted toward BTK, Bcl-2, PKCβ, MALT1, and IKZF1/3. Combination trials are assessing PI3Kδ plus BTK, Bcl-2 plus lenalidomide and rituximab, CDK9 plus Bcl-2, anti-CD32 plus rituximab, anti-CD47 plus rituximab, and vaccine plus lenalidomide and rituximab. A small number of BTK-degrading compounds are also showing promise, but more research is needed to understand safety and dosing. In the future, Stathis expects to see more patients with MZL in phase 1 trials (and expansion cohorts) and would like to see patients with relapsed disease in trials to test new drugs, opportunities to test new drugs once safety data emerge from other studies, incorporation of molecular testing and liquid biopsies, and international efforts toward MZL trials.

Session V: Epidemiology Prognosis and Transformation

To open the second day of the workshop, James R. Cerhan, MD, PhD, professor of epidemiology at the Mayo Clinic College of Medicine and Science, Ralph S. and Beverly Caulkins Professor of Cancer Research, coleader of the Genetic Epidemiology and Risk Assessment Program in the Mayo Clinic Comprehensive Cancer Center, codirector of the Biorepositories Program in the Mayo Clinic Center for Individualized Medicine, and associate director of the Mayo Clinic Cancer Registry, provided an update on the epidemiology of MZL. The incidence of MZL has been increasing since 2001; it increases with age and is higher in men for most subtypes, and in most cases, it is highest in patients who are White than in those in other demographics. The incidences of stomach and salivary gland MZL appear to be decreasing, while skin and lung MZL are increasing, especially in women and patients younger than 50 years old. Five-year survival rates continue to increase and are highest for extranodal MZL at 96%, followed by splenic (85%) and nodal (85%) disease. Established risk factors include infections, autoimmune disease, solid organ transplantation, family history, and certain genetic loci. Potential risk factors include smoking, alcohol use, sun exposure, hair dye, and some occupations. MZL does not appear to cluster with any other non-Hodgkin lymphoma subtypes. New risk factors have been evaluated since the last MZL workshop; updated data did not link lymphoma to glyphosate use, body mass index did not impact risk in adults or young adults, physical activity lowered risk for MZL, and low-dose aspirin was protective against MZL. Emerging data will expand understanding of the epidemiologic patterns of this disease and its subtypes. Cause-of-death analysis studies are aiding in the further understanding of MZL.⁸

Next, Luca Arcaini, MD, professor of hematology at the University of Pavia, discussed the features of transformed MZL (tMZL). MZL can undergo transformation to large B-cell lymphoma, DLBCL, and high-grade large B-cell lymphoma after diagnosis; the identification of tMZL requires histologic assessment. The features of tMZL include a rise in lactate dehydrogenase (LDH) level, hypercalcemia, a sudden decline in performance status, rapid localized nodal growth, new/unexpected extranodal sites of disease, and the presence of new B symptoms. The histology of tMZL is generally straightforward, though some cases show borderline features. Transformation to DLBCL is indicated by the presence of confluent sheets of blasts, but it can be difficult to identify blasts, and the “sheet of blasts” is not clearly defined. tMZL is generally related to a complex karyotype and mutations in NOTCH3, TP53, and TBL1XR1. Of note, Arcaini remarked that a subset of DLBCLs diagnosed as de novo may be tMZL. Current research limitations in tMZL include heterogeneous data quality and series with incomplete or missing data and short follow-up times, which translates to heterogeneous data in incidence rates, risk factors, and outcomes. Estimates of the incidence of transformation are around 3% to 15% at 5 years and 5% to 18% at 10 years; rates seem to vary by disease subtype.^8-11 tMZL is tied to increased mortality and poor survival regardless of subtype.^8,12,13 Risk factors for transformation at diagnosis and in MZL include unique clinical characteristics, lab values, and biological features. Limited data are available to predict prognosis, but the progression of disease at 24 months (POD24) seems to have some applicability here. Limited information is available to guide treatment selection for tMZL; rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) and anthracycline are treatment options for patients with and without previously untreated MZL, respectively, and CAR T-cell therapy is an emerging treatment option. Acknowledging the need for better definitions in this space, Arcaini introduced a retrospective study that is underway to assess clinical and molecular characteristics, pathology, and outcomes in MZL to fill the knowledge gaps.

Juan Alderuccio, MD, associate professor of medicine, Division of Hematology, University of Miami, followed with a discussion of prognostic models in MZL. Patients with MZL are generally known to have good survival; in an analysis of data from the University of Miami, rates of median PFS were highest at 10.6 years in extranodal MZL, followed by nodal (4.8 years) and splenic (3.9 years) MZL, and median OS was not reached. Several useful measures can help assess prognosis in MZL. Key prognostic factors include monoclonal paraprotein, which is correlated with PFS after frontline therapy in MZL, in patients with extranodal disease, and in patients treated with rituximab and immunotherapy and is also tied to the risk for transformation. The MALT-International Prognostic Index (IPI) model considers age 70 years or older, Ann Arbor stage III/IV, and elevated LDH to predict risk (low, intermediate, or high) for EFS in patients with extranodal MZL and other MZL subtypes.^9-15,16 An analysis of data from patients with extranodal MZL at the University of Miami identified multiple mucosal sites (MMS) as a factor correlated with survival in MZL. The Revised MALT-IPI, which includes MMS, categorizes patients into low (score 0), low-medium (score 1), medium-high (score 2), and high-risk (score 3 or higher) groups. The Revised MALT-IPI score was tracked with the rate of POD24 and transformation events. In splenic MZL, 2 prognostic scores are useful: the Italian Lymphoma Intergroup risk score, which considers hemoglobin level more than 12 g/dL, LDH level higher than normal, and albumin level more than 3.5 g/dL; and the HPLL model, which considers hemoglobin concentration, platelet count, elevated LDH, and the presence of extrahilar lymphadenopathy. For assessing MZL as a single entity in patients in need of systemic therapy, the MZL-IPI score includes LDH, absolute lymphocyte count, hemoglobin, platelets, and nodal MZL or disseminated MZL. The MZL-IPI score categorizes patients into low-, intermediate-, and high-risk groups and is correlated with PFS and OS in patients with MZL.¹⁷ Alderuccio also noted that POD24 and failure to achieve CR after frontline therapy are 2 separate factors linked to outcomes in MZL that may have some applications to determining prognosis. Future research will inform the utility of these models in the context of emerging therapies.

Session VI: Assessment Criteria, Response Evaluation, and Surrogate End Points

To kick off this session, Alderuccio returned with an overview of the clinical utility of PET/CT imaging. The use of PET/CT imaging to characterize MZL has been controversial due to the variability in 18fludeoxyglucose (FDG) avidity across the different subtypes and the risk of high tissue avidity obscuring critical information; however, modern technology may be overcoming these hurdles. In an analysis of data from the University of Miami, researchers sought to identify cases of MZL with FDG-avid disease. Across 187 locations in 152 patients, FDG-avid disease was detectable in 78.1% and was detectable across a variety of tumor locations. FDG avidity increased with increasing tumor size, with tumors smaller than 0.5 cm being non–FDG-avid. In patients with multiple mucosal sites, over 80% of patients showed FDG avidity across all disease locations. Not much data exist to describe the role of PET/CT in nodal MZL, and PET/CT showed low sensitivity for detecting bone marrow involvement in MZL. The value of measuring metabolic tumor volume is not clear for MZL and is a subject for future studies. CXCR4 tracers can aid in the detection of MZL via PET/CT but are limited by high rates of splenic tracer uptake and retention. Alderuccio concluded that PET/CT should be included in the staging workup for patients with MZL, but it is important to correlate the findings with other disease information, especially in locations with the potential for obfuscation. Future studies should further characterize the role of PET/CT in response assessment and as part of screening for clinical trials.

Catherine Thieblemont, MD, PhD, head of the Hemato-Oncology Department at the Hôpital Saint-Louis, Paris, France, followed with a discussion of the clinical utility of minimal residual disease (MRD). Defined as the presence of residual cancer cells after treatment in patients with clinically undetectable disease, MRD may be helpful in identifying appropriate treatment and adapting treatment approaches in MZL. Thieblemont discussed how the identification of MRD in patients can be performed using imaging, biological approaches, or a combination of the 2. Of note, in splenic disease, MRD is particularly challenging to implement because it may not exhibit any blood involvement. The role of imaging-detected MRD in MZL is an active area of research; clinical trials are assessing the roles of PET in prognosis and response assessment in MZL, and CT-identified MRD has been used to help guide de-escalation of treatment. Biologic evaluation of MRD (including assessments via measurement of IgH rearrangements and multicolor flow cytometry) has also been used in clinical trials and has been linked to outcomes including response, PFS, OS, and POD24 positivity. The addition of MRD can also improve efficacy assessments by providing more information about how a patient’s disease is responding to treatment.¹⁷ With the emergence of new therapies that lead to deeper responses, MRD may be a useful parameter to compare efficacy between similar treatments. MRD may also have applications as a surrogate end point, with additional data presented in the next session. Thieblemont concluded that MRD can improve clinical assessments in routine practice and has the potential to accelerate the rate at which useful information is obtained from clinical trials. Future research will validate the use of MRD and explore its role in aiding treatment decisions in MZL.

Next, Côme Bommier, MD, of Hôpitaux de Paris and Mayo Clinic, discussed the clinical utility of early surrogate end points. A challenge facing drug development in MZL is the many years required to assess outcomes in clinical trials. In an effort to accelerate the assessment of potential therapies to spare time and resources, surrogate end points are being investigated that more rapidly provide sufficient information about treatment efficacy without sparing patient safety. Only 1 study has shown positive results for a surrogate end point assessed in MZL: the study assessed early CR as a surrogate end point in the phase 3IELSG19 trial (NCT00210353).¹⁸CR was chosen based on data indicating that more patients who were treated with double therapy had a higher rate of and spent more time in CR than those treated with single therapy. CR at 24 months (CR24) and time to CR censored at 24 months were compared with POD24, and results indicated that time to CR and CR24 captured 95% of the information describing 8-year PFS. Using this surrogate end point, the researchers were able to obtain a single measurement at 2 years that provided a strong prediction of 8-year PFS. Bommier concluded that more phase 3 trials are needed in MZL, and further research is needed to fully understand the role of CR24 in MZL.

Session VII: Clinical Trials

To kick off the discussion of clinical trials, Izidore S. Lossos, MD, professor of medicine, chief of the Lymphoma Section at the Division of Hematology, endowed director of the Lymphoma Program, and head of Lymphoma Site Disease Group at the University of Miami Sylvester Comprehensive Cancer Center, provided an overview of the clinical trials in North America (Table). Lossos began with an overview of research needs in MZL treatment, for which there is no established standardized approach. Patients have lengthy survival and some do not require treatment, but the criteria for treatment initiation are based on those for FL and are not applicable to many patients with extranodal and splenic MZL. Only 1 randomized trial has been performed in the rituximab era in patients with extranodal MZL, so treatment guidelines are based on large institutional experiences.

Table 1

Treatment for nodal disease is similar to that for FL, and first-line rituximab is generally used to treat splenic disease. Past trials assessing up-front chemotherapy showed variable responses, and trials in relapsed/refractory MZL showed objective response rates in the 60% to 70% range and variable CR rates, but some potentially useful therapies have been taken out of circulation. Current clinical gaps include the fact that MZL remains incurable for most patients, there are few trials specifically targeting MZL, the role of PET is not understood, inclusion and response criteria are needed, and more efficient treatments are required. No trials are currently recruiting for newly diagnosed MZL, 5 are recruiting to assess treatments in newly diagnosed FL/MZL/low-grade non-Hodgkin lymphoma, 1 is recruiting for recurrent MZL, and 2 are recruiting for FL/MZL/low-grade non-Hodgkin lymphoma. Recruitment has been completed for 2 trials in newly diagnosed MZL and 1 trial in relapsed MZL. Historically, MZL enrollment in clinical trials has been low due to restrictive inclusion criteria–based Lugano recommendations and the lower incidence of MZL. Lossos noted that specific staging and response assessment criteria are needed for most extranodal MZL sites, and he proposed new criteria for assessing treatment response and for the inclusion of patients with MZL in clinical trials. Lossos concluded by highlighting 3 ongoing phase 2 clinical trials in patients with MZL at the University of Miami.

Christian Buske, MD, medical director at the Comprehensive Cancer Center and the Institute of Experimental Cancer Research at Ulm University, Germany, and attending physician and professor of medicine at the Medical Department for Internal Medicine III, Hematology/Oncology, Ulm University Hospital, followed with an overview of European MZL clinical trials. Buske began by introducing the phase 3 GALLIUM study (NCT01332968), an open-label, randomized trial assessing obinutuzumab vs rituximab in a population including adults with previously untreated CD20-positive MZL.¹⁹ The trial had an unacceptable rate of adverse effects, supporting the push toward chemotherapy-free treatment approaches in MZL. Several ongoing studies are assessing chemotherapy-free treatments including BTK inhibitors (phase 2 IELSG47/MALIBU), PI3K inhibitors (phase 2 COUP-1 [NCT03474744]), checkpoint inhibitors (phase 2 POLE-1), and anti-CD20 antibodies (phase 2 OLYMP-1 [NCT03322865]). Planned studies include the phase 3 IELSG48, MAGNOLIA, and MARSUN studies and the phase 2 EPOS-1 study. Buske closed the discussion by sharing 2 academic projects. The first was related to the need for real-world data; the MZL registry is a prospective and academically funded web-based registry that is actively recruiting all adult MZL patients. Second, Buske described an ongoing international retrospective study that aims to identify clinically relevant subgroups of patients with nodal MZL. Patient data and tissue samples from more than 50 international sites are being collected and analyzed.

Future Road Map Toward Progress in MZL: A Roundtable Discussion and Workshop Summary

In the final session of the workshop, Habermann and Rossi led a roundtable discussion of the progress made on the objectives set in the 2019 meeting. The group discussed remaining concerns and unanswered questions, divided into 6 major subtopics, and set short- and long-term goals for addressing the lingering issues.

Biology and Pathology

Concerns and unanswered questions in biology and pathology include understanding differences in the MZL subtypes, the tumor microenvironment, molecular clusters, and immunology related to MZL. Solutions proposed for these knowledge gaps include performing further research to define MZL disease subtypes, exploring the MZL tumor microenvironment, taking an unbiased approach to characterize extranodal disease, and investigating the switch to antigenic stimulation in MZL in the short term. In the long term, participants seek to establish MZL-specific biologic correlates for pathology and diagnostic uses, identify ways to differentiate MZL from LPL and define the source and precursors for MZL.

Epidemiology Transformation

In epidemiology and transformation, current unmet needs include the lack of understanding of the epidemiology of nodal disease, extranodal subsets, and splenic MZL; the lack of consensus around the pathologic diagnosis of transformation; and the need to identify transformation risk, natural history, and underlying biological mechanisms. Aims for the immediate future include efforts to define risk factors for each MZL subtype, harmonize the classification of disease subtypes, and understand the risks and biology related to transformation. In the longer term, the group aims to define the predictors and markers of MZL transformation.

Assessment Criteria, Response Evaluation, and Surrogate End Points in MZL

Persistent issues in disease assessment include the need for improved assessment criteria, response criteria, and end points specific to MZL. Tasks to address these issues in the near future include developing new response criteria including specific response criteria for splenic MZL, understanding the role of MRD and PET and other staging modalities, and identifying the role of end points including minor response, CR, and potential surrogate end points in MZL. In the long term, the participants prioritized the development of novel genomic and radiologic assessments of response.

MZL Targeted Pathways

There also remains a need to improve the field’s knowledge of targetable pathways for MZL treatment, as drug development is not keeping up with what is known about MZL biology. Immediate solutions to this issue are to push to identify new targetable pathways; optimize the applications of CAR T-cell therapy; generate preclinical data; and, in the long term, implement new trials with targeted therapies in the relapsed/refractory setting. The participants emphasized the importance of working with policy makers toward these goals.

Concerns and unanswered questions

• Druggable pathways and preclinical data are limited (MYD88, BRAF, NOTCH, CREBBP, NFkB, p53, FAS [cutaneous]).
• Tumor targets are poorly understood: which drugs and which targets

Immediate action solutions (1-5 years)

• Identify new pathways and targets.
• Target microenvironment.
• Develop new preclinical data.
• Prognostic models integrated with plasmacytic differentiation, histology, genetics, and clinical data are needed.

Long-term solutions (more than 5 years)

• Implement new trials that are targeted therapies in the relapsed/refractory setting.

Etiology and Natural History of MZL Subtypes

The etiology and natural history of MZL subtypes remain inconsistently defined, and there appear to be geographic differences influencing disease biology. To resolve the inconsistencies, the participants set aims to further define the MZL subtypes and select patients within those subsets for treatment or clinical trial enrollment and to define cure in MZL subtypes. An overarching goal is to determine how many different diseases are present within the MZL umbrella.

Concerns and unanswered questions

• The etiology and natural history of MZL subtypes are not uniformly defined.
• There appear to be geographic differences.
• The microbiome needs to be further studied.

Immediate action solutions (1-5 years)

1. Further define MZL lymphoma subtypes with larger data sets.
2. Define which patients in each disease subset need treatment.
3. The role of infectious agents needs further exploration.
4. Define cure in the MZL subtypes.
5. Define appropriate patients for clinical trials in MZL subtypes.

Long-term solution (more than 5 years)

• Determine how many diseases MZL represents.
• Further understanding relationships of MZL disease with age.

Treatment of MZL

Finally, many challenges persist with treating MZL. Treatment patterns are not standardized and may vary widely. Clinical trials typically do not discriminate between patients with MZL and other indolent lymphomas and often do not predefine MZL subtypes, and it is not clear whether local control vs long-term control should be the treatment priority. In the short term, initiatives to address these issues include further research into the MZL subtypes; defining new clinical trial strategies, including those with a focus on MZL; and pursuing orphan disease designation for MZL subtypes. In the long term, participants agreed it would be important to define standards of care for each of the MZL subtypes and to develop curative approaches tailored to each subtype.

Concerns and unanswered questions

• The treatment patterns vary and could be further standardized in MZL.
• Clinical trials routinely include groups of indolent FLs and MZL, and not patients with MZL only.
• MZL subtypes (extranodal, splenic, nodal) are not predefined in clinical trials.
• A major treatment issue is local control vs long-term control.
• Orphan definition needs to be pursued.

Immediate action solutions (1-5 years)

• Further define areas of research for all subtypes of MZL.
• Define new clinical trial strategies.
• Clinical trials should be designed for MZL in certain study designs.
• Trials should define individual subsets.
• Orphan disease designation for individual subsets.

Long-term solutions (more than 5 years)

• Define standards of care for each of the subtypes to benchmark new therapeutic approaches.
• Develop curative approaches to all subtypes of MZL.
• Work with pharmaceutical companies on strategies for different groups of patients (older, eligible…).
• The science is further ahead of treatment.

Summary

The 2024 MZL Scientific Workshop brought together a cohort of experts to discuss recent advancements in MZL biology, diagnosis, characterization, and treatment. This forum provided a platform for the discussion of the state of the field and allowed MZL experts to reflect on recent learnings, identify gaps in knowledge, and develop priorities and strategies to continue to propel the field’s understanding of MZL. Exciting progress has been made since 2019, but continuing efforts are needed to understand and characterize MZL, especially the disease subtypes, to inform diagnosis, assessment, clinical trial design, and ultimately treatment.

Acknowledgments

This meeting, in addition to several projects presented, was supported by grants from the Lymphoma Research Foundation. The steering committee expresses appreciation to the attendees, ADC Therapeutics, and the Foundation staff who participated in fruitful discussion and made this year’s meeting possible. Each presenter whose work is included herein reviewed and approved the summary of that work.

Steering Committee

Thomas Habermann, MD (cochair)

David Rosse, MD (cochair)

Emanuele Zucca, MD

Ming-Qing Du, MD, PhD

Ari Melnick, MD, PhD

Margaret Shipp, MD

Catherine Thieblemont, MD

Loretta Nastoupil, MD

Kami Maddocks, MD

Andrew Zelenetz, MD, PhD

Kojo Elenitoba-Johnson, MD

Izidore Lossos, MD

References:

1. Treon SP, Xu L, Yang G, et al. MYD88 L265P somatic mutation in Waldenström’s macroglobulinemia. N Engl J Med. 2012;367(9):826-833. doi:10.1056/NEJMoa1200710
2. Hopper MA, Wenzl K, Hartert KT, et al. Molecular classification and identification of an aggressive signature in low-grade B-cell lymphomas. Hematol Oncol. 2023;41(4):644-654. doi:10.1002/hon.3187
3. de Martel C, Georges D, Bray F, Ferlay J, Clifford GM. Global burden of cancer attributable to infections in 2018: a worldwide incidence analysis. Lancet Glob Health. 2020;8(2):e180-e190. doi:10.1016/s2214-109x(19)30488-7
4. Chihara D, Huang EP, Finnigan SR, et al. Trends in grade 5 toxicity and response in phase I trials in hematologic malignancy: 20-year experience from the cancer therapy evaluation program at the National Cancer Institute. J Clin Oncol. 2022;40(17):1949-1957. doi:10.1200/jco.21.02190
5. Griguolo G, Zorzi MF, Pirosa MC, et al. A systematic review of contemporary phase I trials in patients with lymphoma. Crit Rev Oncol Hematol. 2022;180:103860. doi:10.1016/j.critrevonc.2022.103860
6. Leonard JP, Trneny M, Izutsu K, et al. AUGMENT: a phase III study of lenalidomide plus rituximab versus placebo plus rituximab in relapsed or refractory indolent lymphoma. J Clin Oncol. 2019;37(14):1188-1199. doi:10.1200/jco.19.00010
7. Opat S, Tedeschi A, Linton K, et al. The MAGNOLIA trial: zanubrutinib, a next-generation Bruton tyrosine kinase inhibitor, demonstrates safety and efficacy in relapsed/refractory marginal zone lymphoma. Clin Cancer Res. 2021;27(23):6323-6332. doi:10.1158/1078-0432.Ccr-21-1704
8. Tun AM, Khurana A, Mwangi R, et al. Causes of death in low-grade B-cell lymphomas in the rituximab era: a prospective cohort study. Blood Adv. 2022;6(17):5210-5221. doi:10.1182/bloodadvances.2022007990
9. Alderuccio JP, Zhao W, Desai A, et al. Risk factors for transformation to higher-grade lymphoma and its impact on survival in a large cohort of patients with marginal zone lymphoma from a single institution. J Clin Oncol. 2018:JCO1800138. doi:10.1200/jco.18.00138
10. Kalashnikov I, Tanskanen T, Viisanen L, Malila N, Jyrkkiö S, Leppä S. Transformation and survival in marginal zone lymphoma: a Finnish nationwide population-based study. Blood Cancer J. 2023;13(1):62. doi:10.1038/s41408-023-00831-9
11. Bommier C, Link BK, Gysbers B, et al. Cumulative incidence, predictors, and outcomes of transformation to diffuse large B-cell lymphoma in a prospective cohort of patients with marginal zone lymphoma. Blood. 2023;142(suppl 1):1672. doi:10.1182/blood-2023-186626
12. Florindez JA, Alderuccio JP, Reis IM, Lossos IS. Splenic marginal zone lymphoma: a US population-based survival analysis (1999-2016). Cancer. 2020;126(21):4706-4716. doi:10.1002/cncr.33117
13. Bult JAA, Huisman F, Zhong Y, et al. A population-based study of transformed marginal zone lymphoma: identifying outcome-related characteristics. Blood Cancer J. 2023;13(1):130. doi:10.1038/s41408-023-00903-w
14. Du Y, Wang Y, Li Q, et al. Transformation to diffuse large B-cell lymphoma and its impact on survival in patients with marginal zone lymphoma: a population-based study. Int J Cancer. 2024;154(6):969-978. doi:10.1002/ijc.34773
15. Thieblemont C, Cascione L, Conconi A, et al. A MALT lymphoma prognostic index. Blood. 2017;130(12):1409-1417. doi:10.1182/blood-2017-03-771915
16. Tracy SI, Larson MC, Feldman AL, et al. The utility of prognostic indices, early events, and histological subtypes on predicting outcomes in non-follicular indolent B-cell lymphomas. Am J Hematol. 2019;94(6):658-666. doi:10.1002/ajh.25473
17. Arcaini L, Bommier C, Alderuccio JP, et al. Marginal zone lymphoma international prognostic index: a unifying prognostic index for marginal zone lymphomas requiring systemic treatment. EClinicalMedicine. 2024;72:102592. doi:10.1016/j.eclinm.2024.102592
18. Lyu R, Wang T, Wang Y, et al. Undetectable minimal residual disease is an independent prognostic factor in splenic marginal zone lymphoma. Br J Haematol. 2021;194(5):862-869. doi:10.1111/bjh.17703
19. Bommier C, Zucca E, Chevret S, et al. Early complete response as a validated surrogate marker in extranodal marginal zone lymphoma systemic therapy. Blood. 2024;143(5):422-428. doi:10.1182/blood.2023020984
20. Herold M, Hoster E, Janssens A, et al. Immunochemotherapy and maintenance with obinutuzumab or rituximab in patients with previously untreated marginal zone lymphoma in the randomized GALLIUM trial. Hemasphere. 2022;6(3):e699. doi:10.1097/hs9.0000000000000699