Lymphoid malignancies comprise a heterogeneous group of leukemias and lymphomas originating from the lymphoid organs, which form a major part of the immune system. Primary lymphoid organs consist of the bone marrow (BM) and thymus, while secondary lymphoid organs include the lymph nodes (LNs) and spleen. Lymphoid cancers are grouped into five main categories: Hodgkin lymphoma, non-Hodgkin lymphoma (NHL), multiple myeloma (MM), and acute and chronic leukemias. However, recent advances in our understanding of the genetic drivers of lymphoid cancers have resulted in a better subclassification of these tumors that differ widely in their phenotype, biology, and clinical behavior [1, 2]. For instance, approximately 20–30% of patients with diffuse large B-cell lymphoma (DLBCL), which is the commonest subtype of NHL, are unable to achieve complete remission with the standard chemoimmunotherapy regimen of rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisolone (R-CHOP) [3]. Furthermore, patients with high-grade B-cell lymphoma (HGBL) with concurrent MYC plus BCL2 and/or BCL6 rearrangements, the so-called double-hit or triple-hit lymphomas, constitute an additional high-risk group with a complete remission (CR) rate of less than 60% with the conventional R-CHOP therapy [4].

The success of the anti-CD20 monoclonal antibody rituximab, which was approved by the Food and Drug Administration (FDA) in 1997 for B-cell lymphoma, highlights the importance of immunotherapy for B-cell malignancies. Rituximab was one of the first form of immunotherapy for cancer treatment, and its mechanisms of action included direct induction of apoptosis, complement-dependent cytotoxicity (CDC), and antibody-dependent cell-mediated cytotoxicity (ADCC) [5], although their relative contributions to its overall effects remain unclear. These effects are mainly mediated by cytotoxic lymphocytes, including CD8⁺ cytotoxic T-cells and natural killer (NK) cells, which play major roles in adaptive and innate immune responses, respectively [6]. T cells can recognize neoantigens generated by cancer-causing mutations or tumor-associated antigens (TAAs), which are highly expressed in tumors but can also be expressed in healthy tissues. Despite the presence of endogenous T-cells in the tumor microenvironment, most tumors progress through the hyporesponsive state of CD8⁺ cytotoxic T-cells or NK cells owing to self-tolerance, which is the host immune mechanism for preventing autoimmunity [7].

CD8⁺ cytotoxic T-cells can detect and directly eradicate tumor cells in an antigen-specific manner. T-cell activation requires three signals: stimulatory signals mainly mediated by T-cell receptor (TCR) complexes (signal 1); co-stimulatory signals controlled by antigen-presenting cells (APCs), such as dendritic cells (signal 2); and cytokines (signal 3). To initiate an immune response, antigens must be captured and presented on T-cells by APCs. Subsequently, the TCR on T-cells recognizes neoantigens or TAA on major histocompatibility complex (MHC) molecules in collaboration with CD8 or CD4 co-receptors. Class I MHC molecules are expressed on all nucleated cells, whereas class II MHC molecules are primarily expressed on APCs and a few other cell types, including thymic epithelial and endothelial cells [8]. MHC molecules require the formation of a complex with peptides for their stable expression on the cell surface as an MHC-peptide complex; Class I MHCs can accommodate 8–11 peptides, while Class II MHCs can accommodate 10–30 or more peptides [9]. Therefore, neoantigens or TAAs must be processed for MHC. However, before their presentation to T-cells, endogenous antigens, such as TAAs are processed by proteasomes for class I MHC molecules, whereas extracellular antigens are ingested and degraded in lysosomes to form complexes with class II MHC molecules [9]. Additionally, TCRs recognize only two or three amino acid residues of a peptide within the MHC-peptide complex as antigenic determinants [9]. TCRs bind only to specific peptide-MHC complexes and not to other molecules; therefore, TCR binding is MHC-restricted. Indeed, T-cells can only recognize cell-associated antigens but not soluble or cell-free antigens.

Overall, characteristics of TCR binding to an MHC-peptide complex includes low affinity, slow kinetics, and high cross-reactivity. This is different from the antigen binding of immuno-globulins of the B-cell receptor (BCR), which can bind to linear and conformational determinants of various target macromolecules, such as proteins, lipids, polysaccharides, and even small chemicals [10]. Additionally, BCR can recognize soluble antigens and bind to antigens with high affinity at a rapid on-rate, unlike TCRs which cannot undergo affinity maturation [10].

CARs are synthetic fusion receptors that redirect T-cells to TAAs, such as CD19 on B-cell lineage tumors or B-cell maturation antigen (BCMA) on MM cells [11]. Most CAR T-cells have features of BCRs in the form of a single-chain variable fragment from the antibody as an antigen recognition domain, which is linked to the intracellular costimulatory domain, such as CD28 or CD137 (4-1BB) [11]. However, the antigen-binding domain of ciltacabtagene autoleucel, a BCMA targeting CAR T-cells for MM, comprises only heavy-chain variable domain without the light chain (nanobody), which may reduce the immunogenicity associated with the linker region [12]. CAR T-cells can recognize TAA independently without the MHC-TCR complex (signal 1), and they have their own co-stimulatory domain (signal 2) that are attached to the activating domain, typically the zeta chain of the CD3 complex, which form the backbone of the current second-generation CAR T-cells [13]. Ever since CAR T-cells were approved for B-cell malignancies in 2017, their unprecedented success in clinical trials in highly refractory patients with lymphoid cancers has resulted in FDA approval of six CAR T-cells against relapsed or refractory subsets of B-cell lymphoma/leukemia and MM (Table 1).

Axicabtagene ciloleucel (axi-cel) and brexucabtagene autoleucel (brexu-cel) have a CD28 costimulatory domain, whereas other FDA-approved CAR T-cells are based on the 4-1BB costimulatory domain. Mechanistically, CD28-based CAR T-cells can elicit a robust proliferative response and yield effector memory T-cells, whereas 4-1BB can lead to a progressive response with enhanced persistence and central memory differentiation [14]. Indeed, 4-1BB CAR T-cells have demonstrated distinct exhaustion features driven by the transcription factor FOXO3 [15] that differ from the classic exhaustion features that can occur in CD28-based CAR T-cells. It is clinically challenging to compare the efficacies of different types of CAR T-cells. However, the French DESCAR-T registry study demonstrated better outcomes with axi-cel than with tisagenlecleucel in terms of response rate and survival [16]. In contrast, real-world studies from the US and Germany have demonstrated no statistically significant differences in the response rates and survival between the two CAR T-cell products [17, 18]. Collectively, it would be valuable to further investigate the issue of selecting CAR T-cells based on different co-stimulation and exhaustion mechanisms to improve CAR T-cell function.

Recent long-term follow-up data suggest that patients with B-cell lymphoma have a lower CR rate than patients with MM following CAR T-cell treatment, although a direct comparison is difficult [19]. Patients with B-cell lymphoma tend to have more durable responses once CR is achieved [19]. Patients with MM demonstrated a higher CR rate than those with B-cell lymphoma; however, they encountered less sustained remission, even after achieving CR, although patients with deeper initial remission are likely to have sustained responses [19]. Indeed, patients with a low tumor burden and less extramedullary disease are more likely to have better responses to CAR T-cell treatment [20]. Moreover, target antigen escape plays an important role in the durability of response and relapse following CAR T-cell therapy, although its relative contributions have been reported to vary significantly across disease subtypes [21]. In this regard, targeting dual antigens simultaneously with CAR T-cells is being explored in pre-clinical and clinical settings [22-24].

Preclinical and clinical studies have demonstrated that CAR T-cells derived from less-differentiated T-cell subsets, such as stem cell memory T-cells (T_SCM) and central memory T-cells, are associated with better CAR cell expansion and clinical response [25]. T_SCM has been demonstrated to contribute to circulating CAR T-cell pools during long-term persistence, with predominant clonal burst immediately after infusion [26]. Of note, single-cell multi-omics analysis has revealed that decade-persisting CD19 CAR T-cells are CD4⁺ CAR T-cells that express cytotoxic molecules with direct cytotoxic functions upon stimulation with CD19-expressing cells ex vivo [27]. Notably, a higher CD4-to-CD8 ratio in leukapheresis products was associated with better in vivo expansion and durable clinical response of BCMA CAR T-cells [28, 29]. Specifically, a higher percentage of naïve and early memory CD4⁺ T-cells was associated with long-term response in patients who received decabtagene vicleucel [30]. In line with these findings, CAR T-cell expansion with a balanced ratio of CD4⁺ naïve and CD8⁺ T_SCM cells reduced the number of infusion products required to achieve clinical responses, which is now commercially available as lisocabtagene maraleucel [31, 32].

Efforts to improve CAR T-cell outcomes based on T-cell-related factors have led to the development of CAR T-cells as an early line of therapy. Data from the ZUMA-7 trial demonstrated the overall survival benefit of second-line axi-cel in patients with early relapse or refractory large B-cell lymphoma (LBCL) who were eligible for autologous stem cell transplantation (ASCT) [33]. This was the first randomized trial in nearly 30 years to demonstrate a survival benefit with second-line treatment for patients with aggressive lymphoma. Indeed, the second line axi-cel was efficacious in patients with relapsed or refractory LBCL who were not eligible for ASCT [34], which accounts for roughly half of the patients with relapsed or refractory LBCL in real-world clinical practice. Furthermore, axi-cel demonstrated efficacy even in the front-line setting for patients with LBCL with high-risk features, including those with positive PET results after two cycles of standard first-line chemoimmunotherapy, either double- or triple-hit, or with high-intermediate- and high-risk International Prognostic Index (IPI) scores [35]. However, no phase 3 trial has compared CAR T-cells with the standard of care in a first-line setting. In this regard, results of the ZUMA-23 trial, which is assessing first-line axi-cel versus chemoimmunotherapy after one cycle of chemoimmunotherapy in high-risk LBCL are awaited, which could delineate the role of front-line CAR T-cell therapy in LBCL. The KarMMa-3 and CARTITUDE-4 trials demonstrated superior progression-free survival with CAR-T cells compared with standard-of-care combinations in earlier lines of MM treatment [36, 37]. Whether CAR T-cell therapy will replace ASCT may be answered by the CARTITUDE-6 trial in which ciltacabtagene autoleucel is compared with ASCT as a consolidation treatment.

One of the major limitations of CAR T-cell therapy is the manufacturing time, which takes at least several weeks [38]. Common adverse events, such as cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) also remain significant issues with CAR T-cell therapy [39]. Furthermore, increasing the number of CAR T-cell deliveries also led to a limited slot for CAR T-cell generation, which may impair timely CAR T-cell treatment for patients with rapid disease progression [40]. Allogenic CAR T-cells have been actively investigated to overcome these issues but their persistence remains poor [41]. In this regard, bispecific T-cell engagers (BiTE) can act as off-the-shelf platforms for T-cell redirection strategies [42]. BiTE links tumor cells and T-cells by binding to both TAA and CD3, which are components of TCR. They have better safety profiles in terms of precise dose titration and they can be discontinued at any time. This can be advantages over CAR T-cells, which are administered as a single infusion with varying doses as a dividing drug. Both CAR T-cells and BiTEs activate T-cells independently of MHC expression on target cells, but their immunologic synapses are quite different. BiTEs form well-organized immunological synapses between T cells and target cells, similar to conventional T cells for the delivery of lytic granules. However, the immunological synapses of CAR T-cells are disorganized, which may paradoxically promote CAR T-cell degranulation [43].

CAR T-cells have better efficacy when derived from less differentiated T-cell subsets; however, bispecific antibodies mediate T-cell effector functions mainly by redirecting more differentiated effector memory T-cells [44]. Blinatumomab was the first BiTE targeting CD19 and CD3 and is now approved for B-cell acute lymphoblastic leukemia (ALL) in first or second complete remission with minimal residual disease (MRD) ≥0.1% based on a BLAST trial [45]. The outcomes from the MajesTEC-1 and MagnetisMM-3 trials, which evaluated BCMA-targeting bispecific antibodies, teclistamab and elranatamb, respectively, led to the approval of these bispecific antibodies by the FDA for relapsed or refractory MM in the fifth-line setting [46, 47]. Additionally, the MonumenTAL-1 trial demonstrated the efficacy and safety of the bispecific antibody talquetamab, which targets GPRC5D [48], thus leading to its approval as the fifth-line treatment for relapsed or refractory MM. CD20 and CD3 bispecific antibodies—mosunetuzumab, epcoritamab, and glofitamab—have also recently been approved for the treatment of both subtypes of aggressive and indolent B-cell lymphoma [49-51]. Mosunetuzumab, epcoritamab, and glofitamab have demonstrated efficacy in relapsed or refractory DLBCL, follicular lymphoma (FL), transformed FL, primary mediastinal large B-cell lymphoma, mantle cell lymphoma, and Richter’s transformation [49-51]. Notably, glofitamab demonstrated favorable efficacy in terms of duration or response in relapsed or refractory DLBCL as 70% of patients continue to be in CR after 18 months [51]. Therefore, bispecific antibodies represent effective approaches for subsets of relapsed or refractory lymphoid malignancies (Table 2), and bispecific combination approaches targeting various antigens are currently active areas of investigation [52].

Natural killer (NK) cells are innate lymphoid cells that recognize target T-cells in a manner that is not restricted by the MHC-peptide complex [53]. Their effector functions are regulated by a balance of signals of various inhibitory and activating receptors [54]. CARs can also be expressed on NK cells, which are similar to the CAR design for T-cells that including a single-chain variable fragment, a co-stimulatory domain, and CD3 zeta. However, recent studies have attempted to design CAR NK cells by employing activating signals associated with NK cell functions. For example, CD3 zeta may be replaced by DNAX-activation protein 10 (DAP10), which induces NK cytotoxicity via signalling through the activating receptor NKG2D–DAP10 complex [55]. NKG2D is one of the major activating receptors on NK cells that recognizes NKG2D ligands, which are stress-induced and are normally not expressed in healthy cells [56]. Indeed, NKG2D CAR NK cells have demonstrated better efficacy than NKG2D CAR T cells in preclinical models of MM [57]. CAR NK cells have the advantage of being an off-the-shelf platform with various gene editing options to enhance efficacy and they have better safety profile compared with CAR T. However, NKG2D ligands are shed from leukemic cells or MM cells to form soluble NKG2D ligands during progression, which can downregulate NKG2D and impair NK cytotoxicity [58]. Consequently, approaches targeting soluble NKG2D ligands have been developed in solid tumor models and in MM to promote NK cell effector functions [59, 60]. In addition to NKG2D-mediated effector functions, ADCC is an important mechanism of NK cell-mediated tumor killing. Our group and others have demonstrated that a subset of NK cells, known as adaptive NK cells, mediate superior ADCC compared to conventional NK cells against MM cells [61, 62]. Therefore, promoting conventional NK cells with CAR and harnessing the superior ADCC of adaptive NK cells with monoclonal antibodies, while clearing soluble NKG2D ligands, may be an effective combination strategy for promoting NK cell-based immunotherapy outcomes [63].

Cellular immunotherapy using CAR T-cells and BiTEs has emerged as a transformative approach for treating lymphoid malignancies. The remarkable progress made in recent years, in addition to ongoing innovative strategies for harnessing NK cells, will provide new hope for patients with these aggressive diseases. A better understanding of both disease biology and the basic principles of the resistance mechanisms of cytotoxic lymphocytes will enable us to improve the efficacy and safety of cellular immunotherapy associated with improved patient survival. Therefore, collaborations between basic and translational researchers, clinicians, and healthcare providers are pivotal to harnessing the full potential of cellular immunotherapy in lymphoid malignancies.