Axicabtagene ciloleucel (axi-cel) has demonstrated substantial efficacy in R/R large B-cell lymphoma (LBCL), particularly among patients who have failed multiple prior lines of therapy [16]. In the 5-year follow-up of the pivotal ZUMA-1 trial, axi-cel achieved an overall response rate (ORR) of 83% and a complete response (CR) rate of 58% [17]. The median duration of response (DOR) was 11.1 months, and the 5-year overall survival (OS) rate reached 43%, highlighting its potential to induce durable remissions in a subset of patients. Supporting these findings, a multicenter real-world analysis from the U.S. Lymphoma CAR T Consortium reported a comparable 5-year OS of 40% and lymphoma-specific survival of 53% [18]. Notably, nearly one-third of patients remained alive and in remission at 5 years, underscoring the curative potential of axi-cel even outside the clinical trial setting.

Axi-cel has also shown impressive outcomes as a second-line treatment for R/R LBCL [19]. In the phase 3 ZUMA-7 trial, axi-cel significantly outperformed standard of care (SOC)—chemoimmunotherapy followed by high-dose chemotherapy and autologous stem cell transplant (ASCT) in eligible patients. Extended analysis revealed a median event-free survival (EFS) of 8.3 months with axi-cel, compared to 2.0 months for SOC (p < 0.001) [20]. Median OS was not reached in the axi-cel arm versus 31.1 months for SOC (p=0.03), reinforcing its role as a curative-intent option for early-relapsing or refractory disease (Table 1).

In indolent B-cell lymphomas, axi-cel has also demonstrated high efficacy. The phase 2 ZUMA-5 trial in patients with R/R FL reported an ORR of 94% and a CR rate of 79% [21]. With a median follow-up of 41.7 months, the median progression-free survival (PFS) was 40.2 months, reflecting sustained clinical benefit [22]. Updated analyses extended the median PFS to 57.3 months, with over half of patients maintaining response at 5 years, emphasizing the long-term disease control and potentially curative outcomes axi-cel offers in indolent lymphomas [23].

Tisagenlecleucel (tisa-cel) has shown meaningful clinical activity as a third-line or later therapy in patients with R/R LBCL (Table 1). In the pivotal phase 2 JULIET trial, tisa-cel achieved an ORR of 52% and a CR rate of 40% in a heavily pretreated population [24]. The median PFS was 5.9 months, and the 24-month PFS rate was 33%, underscoring the potential for durable remissions in select patients [25]. Tisa-cel was also evaluated as a second-line treatment in the phase 3 BELINDA trial, which compared it with standard salvage therapy followed by high-dose chemotherapy and ASCT in patients with early-relapsing or refractory LBCL [26]. However, the study did not meet its primary endpoint of improved EFS, failing to demonstrate superiority over standard care. These results emphasize the importance of patient selection and timing of CAR T-cell therapy in the second-line setting. In FL, tisa-cel has also shown promising efficacy. The phase 2 ELARA trial reported an ORR of 86% and a CR rate of 69% in patients with R/R FL [27]. Updated results from 2024, with a median follow-up of 29 months, demonstrated a 24-month PFS rate of 57% and an OS rate of 88%, with no new safety signals [28]. Notably, comparisons between ELARA and real-world data showed a 1.9-fold higher CR rate and a 1.4-fold higher 12-month PFS rate with tisa-cel, along with an 80% reduction in the risk of death compared to usual care, reinforcing its clinical benefit in this setting [29].

Lisocabtagene maraleucel (liso-cel) has demonstrated robust clinical efficacy across a range of R/R B-NHLs. In the pivotal TRANSCEND NHL 001 trial, which enrolled heavily pretreated patients with R/R LBCL, liso-cel achieved an ORR of 73% and a CR rate of 53% [30]. The median DOR was 23.1 months, and the median PFS was 6.8 months, with a 2-year OS rate of 50.5%, underscoring its potential for durable remission in this high-risk population [31]. In the second-line setting, the phase 3 TRANSFORM trial evaluated liso-cel in patients with primary refractory LBCL or those who relapsed within 12 months of first-line therapy. The trial demonstrated a significant improvement in EFS with liso-cel compared to SOC therapy, which included high-dose chemotherapy and ASCT. The median EFS was 10.1 months with liso-cel versus 2.3 months with SOC (hazard ratio, 0.34; 95% confidence interval, 0.22 to 0.51; p < 0.0001), establishing liso-cel as a superior option in this setting [32].

In FL, the phase 2 TRANSCEND FL trial reported impressive outcomes in both third- or later-line and second-line settings. In the later-line cohort, liso-cel achieved an ORR of 97% and a CR rate of 94%. In the second-line setting, all responses were complete, yielding a CR rate of 96% [33]. At the 12 months, DOR rates were 82% and 90% in ≥ third-line and second-line groups, respectively, with corresponding PFS rates of 81% and 91%. Updated 2-year data presented at ASH 2024 confirmed sustained benefit, with 24-month PFS rates of 72.5% and 82.6%, respectively, in these cohorts [34,35]. Liso-cel has also shown promising activity in MCL. In the MCL cohort of the TRANSCEND NHL 001 trial, patients who had received at least two prior lines of therapy, including a Bruton tyrosine kinase (BTK) inhibitor, achieved an ORR of 85% and a CR rate of 68%. The median DOR was 13.3 months, suggesting meaningful durability of response in this challenging population [36].

Brexucabtagene autoleucel (brexu-cel) is a CD19-directed CAR T-cell therapy that has demonstrated high efficacy in R/R MCL (Table 1), including patients with prior exposure to BTK inhibitors. The pivotal phase 2 ZUMA-2 trial established its role in this setting, reporting an ORR of 93% and a CR rate of 67% among patients who had received multiple prior therapies, including a BTK inhibitor and anti-CD20 monoclonal antibody [37]. The median DOR was 28.2 months, underscoring its potential to induce long-term remission in a population with historically poor outcomes. Subsequent data from ZUMA-2 cohort 3 evaluated brexu-cel in BTK inhibitor–naïve patients, showing similarly encouraging results: an ORR of 91%, a CR rate of 73%, and a 12-month OS rate of 90% [38]. These findings suggest that brexu-cel may also be effective earlier in the treatment course, expanding its therapeutic utility. Real-world studies have further validated the efficacy of brexu-cel in routine clinical practice. A multicenter analysis conducted by the U.S. Lymphoma CAR T Consortium included patients with high-risk features often excluded from clinical trials. This cohort demonstrated an ORR of 90% and a CR rate of 82%, with 6- and 12-month PFS rates of 69% and 59%, respectively [39]. These real-world data confirm the durability and consistency of brexu-cel’s benefit across a broader, more diverse patient population.

Recent advancements in CD19-directed CAR T-cell therapy have led to the development of several next-generation candidates aimed at improving efficacy, minimizing toxicities, and addressing resistance mechanisms. These novel approaches are redefining the landscape of immunotherapy for B-cell lymphomas beyond currently U.S. Food and Drug Administration–approved treatments. Anbalcabtagene autoleucel (anbal-cel) incorporates dual gene silencing of prog-rammed death-1 (PD-1) and TIGIT—two key immune checkpoint receptors—designed to mitigate T-cell exhaustion and enhance antitumor activity within the immunosuppressive tumor microenvironment (TME). In a phase 2 study, anbal-cel demonstrated promising clinical activity in patients with R/R LBCL, with high response rates and durable remissions [40]. WZTL-002, a third-generation CAR T-cell therapy, integrates both CD28 and Toll-like receptor 2 (TLR2) co-stimulatory domains to augment T-cell activation while potentially reducing cytokine-mediated toxicity. In the phase 1 ENABLE-1 trial, WZTL-002 showed a favorable safety profile, with no grade ≥ 3 CRS and only one case of grade 1 ICANS (3%). Clinical efficacy was also encouraging, with CR rates of 52% in the dose-escalation cohort and 56% at the recommended phase 2 dose. These findings suggest that incorporating TLR2 may support outpatient administration by enhancing safety without compromising therapeutic activity [41]. An innovative CD19 CAR construct, 19(T2)28z1XX, was developed through selective mutagenesis of two out of three immunoreceptor tyrosine-based activation motifs, creating a more controlled activation signal. In a clinical study involving R/R DLBCL, 19(T2)28z1XX therapy achieved an ORR of 88% and a CR rate of 75%, even at the lowest tested dose (25×10⁶ CAR T-cells). Toxicity was minimal, with only 4% of patients experiencing grade ≥ 3 CRS and 7% experiencing ICANS. Responses were durable, with most responders remaining in remission beyond 12 months [42]. Next-generation products such as rapcabtagene autoleucel (YTB323) and GLPG5101 have been designed to preserve early memory T-cell phenotypes while significantly reducing manufacturing times. YTB323 is produced using the T-Charge platform in less than 2 days and demonstrated an ORR of 88% and a CR rate of 65% in a phase 2 study of R/R DLBCL. Importantly, severe CRS and ICANS rates were low at 6% and 3%, respectively [43]. Similarly, GLPG5101, developed using a decentralized 7-day fresh product workflow, showed an ORR of 88% and a CR rate of 83% across multiple B-NHL subtypes in the ongoing ATALANTA-1 trial [44]. Both therapies demonstrated robust in vivo expansion and long-lasting responses, particularly in patients achieving early CR and measurable residual disease negativity. These emerging therapies highlight the evolving potential of CD19 CAR T-cell therapy to deliver more potent, durable, and accessible treatment options for patients with B-cell lymphomas. By addressing key barriers such as immune evasion, toxicity, and manufacturing logistics, next-generation CAR T-cell therapies may further expand their role in routine clinical practice. Continued clinical evaluation will be critical to define optimal strategies for their use and integration into current treatment algorithms.

Allogeneic CD19-directed CAR T-cell therapies are emerging as promising alternatives to autologous products, with the potential to address key challenges such as prolonged manufacturing timelines, inconsistent product quality, and limited scalability. These “off-the-shelf” therapies offer the advantage of immediate availability and standardized manufacturing, and recent clinical trials have demonstrated encouraging efficacy and safety. Cemacabtagene anseqedleucel (cema-cel), formerly ALLO-501A, is an allogeneic CAR T-cell product evaluated in a phase 1 trial involving patients with R/R LBCL who were naïve to CAR T-cell therapy [45]. Cema-cel demonstrated an ORR of 58% and a CR rate of 42%. Among patients achieving a CR, the median DOR was 23.1 months. CAR T-cell expansion was observed post-infusion, with persistence detectable up to four months. Importantly, the therapy showed a favorable safety profile, with no cases of graft-versus-host disease, grade ≥ 3 CRS, or neurotoxicity. CTX112 is a next-generation, CRISPR-Cas9–engineered allogeneic CAR T-cell therapy incorporating multiple targeted gene edits, including disruption of TRAC (T cell receptor alpha constant), B2M (beta-2 microglobulin), TGFBR2 (transforming growth factor beta receptor II), and ZC3H12A (zinc finger CCCH-type containing 12A, encoding Regnase-1) [46]. These modifications are designed to improve immune evasion, enhance antitumor function, and increase cell persistence. Early clinical data revealed an ORR of 67% and a CR rate of 44%, with some responses maintained for over one year. Enhanced expansion and in vivo persistence were attributed to the disruption of TGFBR2 and Regnase-1, key regulators of T-cell exhaustion and immune suppression. CB-010, assessed in the ongoing ANTLER trial, is another allogeneic CAR T-cell product engineered to knock out PD-1 expression, thereby augmenting T-cell effector function [47]. In this trial, CB-010 achieved an ORR of 94% and a CR rate of 69% in patients with R/R B-NHLs, underscoring the therapeutic benefit of immune checkpoint disruption in enhancing CAR T-cell efficacy. Collectively, these investigational therapies illustrate the rapid progress of allogeneic CD19 CAR T-cell platforms. By employing sophisticated gene-editing strategies to optimize cell function, persistence, and safety, these approaches are redefining the therapeutic potential of off-the-shelf CAR T-cell products for B-cell malignancies. Continued clinical validation is essential to determine their long-term efficacy, scalability, and role in future treatment algorithms.

Mosunetuzumab is a humanized IgG1-based CD20×CD3 BsAb that has demonstrated clinical efficacy in patients with B-NHLs, particularly FL (Table 1). In a pivotal single-arm phase 2 trial that enrolled 90 patients with R/R FL, mosunetuzumab was administered intravenously in 21-day cycles with a step-up dosing (SUD) strategy during cycle 1 to mitigate CRS: 1 mg on cycle 1 day 1 (C1D1), 2 mg on C1D8, 60 mg on C1D15 and cycle 2 day 1 (C2D1), followed by 30 mg on day 1 of subsequent cycles [48]. Among patients who had received at least two prior lines of therapy—including anti-CD20 therapy—mosunetuzumab achieved an ORR of 80% and a CR rate of 60%. An updated analysis with a median follow-up of 37.4 months showed a 3-year PFS rate of 43.2% [49]. These results led to the regulatory approval of mosunetuzumab for R/R FL after two or more prior lines of therapy in 2022. More recently, subcutaneous administration of mosunetuzumab as a first-line treatment in patients with high-tumor-burden FL yielded a best ORR of 95% and a CR rate of 80% [50]. At a median follow-up of 10.6 months, the estimated 12-month PFS rate was 88%. The treatment was well tolerated, with no neurotoxicity and only low-grade CRS events reported. In a phase 1/2 trial involving patients with R/R DLBCL, mosunetuzumab monotherapy demonstrated an ORR of 42% and a CR rate of 23.9% [51]. This cohort included heavily pretreated individuals, 29.5% of whom had received prior CAR T-cell therapy. The median PFS was 3.2 months, with a median time to first response of 1.4 months. In an updated analysis, responses remained consistent, with a median DOR of 10.6 months among patients who achieved CR [52].

Epcoritamab is a humanized IgG1-based CD20×CD3 BsAb administered following a SUD schedule of 0.16 mg on day 1 and 0.8 mg on day 8, followed by full-dose 48 mg weekly. In the EPCORE NHL-1 trial, which enrolled 157 patients with R/R LBCL and a median of three prior lines of therapy, patients received a median of five treatment cycles. At a median follow-up of 10.7 months, epcoritamab demonstrated an ORR of 63% and a CR rate of 39% [53,54]. Among patients with prior CAR T-cell therapy, the ORR and CR rate were 54% and 34%, respectively. With extended follow-up (median, 25.1 months), the CR rate remained durable at 40.1%, and the estimated 24-month duration of CR was 64.2% [55]. While the median PFS in the overall population was 4.4 months, patients who achieved CR experienced significantly improved outcomes, with estimated 24-month PFS and OS rates of 65.1% and 78.2%, respectively. In the FL cohort, 128 patients who had received at least two prior lines of therapy—including an anti-CD20 monoclonal antibody and either an alkylating agent or lenalidomide—were treated with epcoritamab. After a median follow-up of 17.4 months, the ORR was 82%, with a CR rate of 62.5% [54]. The median PFS was 15.4 months, while the median DOR, duration of CR, and OS had not yet been reached at the time of analysis.

Odronextamab is an IgG4-based CD20×CD3 BsAb administered at doses ranging from 0.1 to 320 mg weekly for the first 9 weeks, followed by dosing every other week until disease progression. In the phase 1 ELM-1 trial, which included 145 patients with R/R B-cell NHL, clinical activity was observed at doses ≥ 80 mg in patients with DLBCL and ≥ 5 mg in those with FL [56]. The median number of prior therapies was three, and 41% of the 85 DLBCL patients had previously received CAR T-cell therapy. In the DLBCL cohort, ORR were similar between CAR T-exposed and CAR T-naïve patients, with ORRs of 33% and 39%, respectively, and a CR rate of 24% in both groups. Among FL patients, 78% achieved an objective response, including a 63% CR rate. The phase 2 ELM-2 trial evaluated odronextamab in multiple disease-specific cohorts using a revised step-up dosing regimen to mitigate CRS. The regimen began with 0.7 mg split across cycle 1 days 1 and 2, followed by stepwise dose escalation to 160 mg on C2D1. Treatment was continued at 160 mg weekly through cycle 4, then increased to 320 mg every 2 weeks until progression or unacceptable toxicity. In the DLBCL cohort (n=127), odronextamab yielded an ORR of 52% and a CR rate of 31.5%, with a median DOR of 10.2 months and median duration of CR of 17.9 months [57]. A dedicated analysis of the CAR T-exposed DLBCL subgroup from ELM-1 confirmed sustained efficacy, with an ORR of 48.3% and a CR rate of 31.7% in 60 heavily pretreated patients [58]. In the FL cohort of ELM-2, patients received an initial step-up dose (0.7/4/20 mg) during cycle 1, followed by 80 mg on days 1, 8, and 15 of cycles 2 through 4. Among 128 patients with R/R FL, the ORR was 80.5%, with a CR rate of 73.4%. Responses were durable, with a median DOR of 23 months, median duration of CR of 25 months, and a median PFS of 20.7 months. The 2-year PFS rate was 46.1% [59]. In the phase 3 OLYMPIA-1 trial for high-risk FL, odronextamab monotherapy in the safety lead-in cohort resulted in a 100% CR rate among response-evaluable patients at Week 12. No dose-limiting toxicities, nor any grade ≥ 2 CRS or ICANS events, were reported [60]. Additionally, in the marginal zone lymphoma (MZL) cohort of ELM-2, odronextamab achieved an ORR and CR rate of 79.3% [61]. Responses were durable, with a 36-month DOR of 72.4% and a 36-month PFS rate of 69.4%. No cases of ICANS or grade ≥ 3 CRS were observed.

Glofitamab is a T cell–engaging BsAb featuring a unique 2:1 configuration that enables bivalent binding to CD20 and monovalent binding to CD3, enhancing T cell recruitment and cytotoxicity. It has been evaluated in the pivotal, phase 1/2 NP30179 trial in patients with R/R B-NHLs. In the phase 1 dose-escalation portion (0.005-30 mg every 14 or 21 days following obinutuzumab pretreatment), glofitamab demonstrated an ORR of 61% and a CR rate of 49% in 171 patients with B-NHLs [62]. In the phase 2 expansion cohort of 155 patients with R/R LBCL, the ORR was 52%, with 39% achieving CR—most within 42 days of treatment initiation [63]. Notably, CR rates were similar among patients who had relapsed after CAR T-cell therapy (35% CR). The 12-month PFS and OS rates were 37% and 50%, respectively, with durable remissions and rare relapses following CR. In an updated analysis with over three years of follow-up, the median duration of CR was 29.8 months. Among patients who achieved a CR at the end of treatment, the 2-year PFS and OS rates were 57% and 77%, respectively [64]. Circulating tumor DNA analysis confirmed sustained molecular remission, and no new safety signals—including CRS or ICANS—were reported with longer follow-up. In the NP30179 trial cohort of R/R MCL (n=60), glofitamab achieved an ORR of 85.0% and a CR rate of 78.3% [65]. With a median follow-up of 17.2 months, the median duration of CR was 15.4 months, and 59.6% of CRs were ongoing at the data cutoff. The estimated 12-month CR duration and PFS rates were 71.0% and 66.6%, respectively. Particularly in patients refractory to prior BTK inhibitor therapy, the ORR and CR rates were 74.2% and 71.0%, respectively. Glofitamab is currently being further evaluated in the ongoing phase 3 GLOBRYTE study in patients with R/R MCL.

Several emerging CD20×CD3 BsAbs are currently under investigation for their therapeutic potential in B-cell lymphomas. Plamotamab, a humanized BsAb comprising an anti-CD20 Fab and an anti-CD3 scFv with an engineered Fc domain to eliminate FcγR binding, has demonstrated clinical activity in heavily pretreated patients with R/R DLBCL and FL. In a phase 1 trial, plamotamab achieved an ORR of 47.4% and a CR rate of 26.3% in DLBCL, and an ORR of 100% with a CR rate of 50% in FL. Importantly, patients with prior CAR-T therapy also responded, with an ORR of 46.2% [66]. IGM-2323 (imvotamab) is a first-in-class IgM-based BsAb that incorporates ten high-affinity CD20-binding domains into an IgM framework. This unique configuration enhances avidity for target cells, enabling robust T cell–mediated cytotoxicity even in cells with low CD20 expression, while also supporting complement-dependent cytotoxicity. In a phase 1 dose-escalation study involving patients with R/R NHL, IGM-2323 demonstrated a 35% ORR among 23 patients with either indolent or aggressive B-NHLs [67]. Ongoing phase 2 randomized trials in DLBCL and FL are evaluating two dose regimens (arm A: 15/100 mg vs. arm B: 15/300 mg) to identify the optimal therapeutic dose. GB261 is a novel CD20×CD3 BsAb designed to retain Fc effector functions, thereby expanding its mechanisms of tumor cell killing through antibody-dependent cellular cytotoxicity and complement-dependent cytotoxicity. In a phase 1/2 trial of 47 patients with R/R B-NHLs treated with doses ranging from 1 to 300 mg, GB261 achieved an ORR of 73% and a CR rate of 46% [68]. At the 100 mg dose level, all five evaluable patients responded, including 80% with CR. Responses were rapid, with a median time to response of 1.3 months. The pharmacokinetic profile showed an effective half-life of 2-3 weeks, supporting dosing intervals of every 3-4 weeks. These results highlight its potential efficacy in heavily pretreated populations. MBS303, a CD20×CD3 BsAb with a 2:1 binding configuration, has also shown encouraging safety and efficacy in a phase 1 dose-escalation study [69]. Among patients with R/R B-NHLs receiving doses ≥ 9 mg, the ORR was 71% with a CR rate of 43%. Notably, in the FL subgroup treated at 9 mg, both the ORR and CR rate reached 100%.

CD19, a widely expressed pan-B-cell antigen, is the target of all currently approved CAR T-cell therapies for B-cell NHL, as well as agents like tafasitamab and loncastuximab tesirine for R/R DLBCL [70,71]. However, CD19 expression is more variable in malignant cells than CD20, contributing to antigen loss in ~30% of patients who relapse after CAR T-cell therapy [72]. Recently developed CD19×CD3 BsAbs show encouraging early efficacy and favorable safety profiles, potentially filling treatment gaps in R/R B-cell lymphoma, including cases previously treated with CD19- or CD20-directed therapies.

In a first-in-human phase 1 trial, AZD0486—a novel fully human CD19×CD3 BsAb engineered for low-affinity CD3 binding to reduce cytokine release—showed promising efficacy and tolerability in R/R DLBCL, including CAR T-exposed patients [73]. Among 40 patients treated at doses ≥ 2.4 mg, the ORR was 43% with a CR rate of 33%. In CAR T-naïve patients receiving 7.2 mg, ORR and CR reached 80%; in CAR T-exposed patients, 36% and 29%, respectively. Treatment was well tolerated, with manageable and reversible CRS and ICANS. Another investigational BsAb, RO7227166, targets CD19 and 4-1BB and is under phase 1 evaluation in combination with glofitamab [74]. Preliminary results from 56 efficacy-evaluable patients with R/R DLBCL showed an ORR of 67% and CR rate of 39%, without new safety concerns.

CRS is a common and significant adverse effect associated with T-cell–redirecting therapies used in the treatment of B-cell lymphoma [75,76]. It arises from rapid T-cell activation and subsequent cytokine release, leading to systemic inflammation. Clinically, CRS can range from mild symptoms such as fever and fatigue to life-threatening complications including hypotension and multi-organ failure. Effective recognition and management of CRS are critical to ensuring the safe and effective use of these therapies.

In phase 3 trials of CAR T-cell therapy as a second-line treatment for R/R LBCL, axi-cel induced CRS in 92% of patients, primarily grade 1-2, with 6% experiencing grade ≥ 3 events [19]. In contrast, tisa-cel caused CRS in 61.3% of patients in the BELINDA trial, with 5.2% experiencing grade ≥ 3 events [26]. The TRANSFORM trial showed that liso-cel led to CRS in 49% of patients, mostly grade 1-2, with only 1% experiencing grade 3 events [32]. Similarly, in the phase 2 PILOT study of liso-cel in transplant-ineligible R/R LBCL patients, grade 3 CRS occurred in only 1.6% of patients (n=61) (Table 2) [77]. For third- or later-line treatment of LBCL, the ZUMA-1 trial reported that axi-cel induced CRS in 93% of patients, again mostly grade 1-2 (37% grade 1, 44% grade 2), with 13% experiencing grade ≥ 3 events [16]. In the JULIET trial, tisa-cel was associated with CRS in 58% of patients, including 36% with grade 1-2 and 22% with grade 3-4 events [24]. The TRANSCEND NHL-001 study showed that liso-cel led to CRS in 42% of patients, with the majority being grade 1-2 (40%) and only 2% experiencing grade 3-4 CRS [30].

In R/R FL, CRS is a common adverse event with CAR T-cell therapy, though its incidence and severity vary by product. In ZUMA-5, axi-cel induced CRS in 78% of patients, predominantly grade 1-2 (72%), with 6% experiencing grade ≥ 3 events, including one fatal case [21]. The ELARA trial of tisa-cel, which enrolled a more heavily pretreated and high-risk FL population (median of four prior therapies; 62.9% with progression of disease within 24 months), reported CRS in 49% of patients—all grade 1-2 [27]. In the phase 2 TRANSCEND FL study, liso-cel was associated with a CRS rate of 58%, with only 1% of patients experiencing grade 3 events [33]. In R/R MCL, CRS occurred in 91% of patients, with most events being grade 1-2 (76%) and 15% grade 3-4 after brexu-cel therapy in the ZUMA-2 trial [37]. More recently, liso-cel was approved for R/R MCL based on the TRANSCEND NHL 001 study, which showed a markedly lower CRS profile, with only 1% of patients experiencing grade 3-4 CRS [36]. Across various lymphoma subtypes, liso-cel appears to be associated with a consistently lower incidence and severity of CRS compared to axi-cel and tisa-cel. In both LBCL and FL, liso-cel demonstrated reduced rates of CRS and fewer high-grade events. Additionally, patients with FL generally experienced milder CRS across all CAR T-cell therapies, suggesting that disease biology and tumor burden may significantly influence both the incidence and severity of CRS in CAR T-cell treatment.

In the context of BsAbs for R/R LBCL, CRS is a commonly reported adverse event but is generally less frequent and severe than with CAR T-cell therapies. In the phase 2 EPCORE NHL-1 study, epcoritamab induced 49.7% of CRS with most events being low grade (32% grade 1, 15% grade 2, and 2.5% grade 3) [53]. Glofitamab, administered intravenously over 12 fixed-duration cycles following pretreatment with obinutuzumab to mitigate CRS risk, resulted in CRS in 63% of 154 patients. Most events were low grade (47% grade 1, 12% grade 2), with 3% grade 3 and 1% grade 4 events [63]. In the phase 2 ELM-2 study of odronextamab, CRS occurred in 55% of 141 patients, with nearly all events being grade 1-2, except for a single case of grade 3 CRS [78]. Mosunetuzumab was associated with a lower CRS incidence in LBCL, occurring in 26.1% of patients—20.5% grade 1, 3.4% grade 2, and 2.3% grade 3—with no grade 4 or 5 events reported [51].

In R/R FL, mosunetuzumab led to CRS in 44% of 90 patients, mostly grade 1-2 (42%), with grade 3 and 4 events each occurring in 1% of patients [48]. Epcoritamab, approved for continuous therapy in R/R FL, showed a higher CRS rate of 67% among 128 patients, with the vast majority being grade 1-2 (65%) and 2% grade 3 events [54]. In R/R MCL, glofitamab remains the most extensively studied BsAb. In a cohort of 37 patients receiving obinutuzumab pretreatment at two different doses, CRS events were mostly low grade. Notably, grade ≥ 3 CRS occurred less frequently in the cohort pretreated with 2,000 mg of obinutuzumab (9.5%) compared to the 1,000 mg cohort (25%) [79].

CRS with BsAbs typically occurs early in treatment, most commonly during the SUD phase. Compared to CAR T-cell therapies, the incidence and severity of CRS with BsAbs are generally lower [80]. This favorable safety profile is largely attributed to mitigation strategies, including SUD protocols and pretreatment with agents such as obinutuzumab, which help temper the initial immune activation and reduce CRS risk.

Neurotoxicity is a notable adverse event associated with T-cell–mediated therapies, commonly termed ICANS. This complication typically arises within days to weeks following treatment and can range from mild cognitive disturbances to severe manifestations such as cerebral edema and seizures. Although the precise pathophysiology remains unclear, it is thought to involve inflammatory cytokine release, blood-brain barrier disruption, and immune cell infiltration into the central nervous system [81]. In prior phase 3 trials evaluating CAR T-cell therapy as second-line treatment for R/R LBCL, tisa-cel was associated with neurological complications in 10.3% of patients, with only 1.9% experiencing grade ≥ 3 events [26]. Liso-cel had a 12% incidence of neurotoxicity, which was mostly low grade (grade 1: 5%, grade 2: 2%, grade 3: 4%) [32]. In the phase 2 PILOT study involving transplant-ineligible patients, liso-cel showed a low incidence of neurologic events (4.9%) and no grade 4 toxicities among 61 treated patients [77]. In contrast, axi-cel demonstrated a significantly higher rate of neurotoxicity. In the ZUMA-7 trial, 60% of patients experienced neurologic events, with 21% being grade ≥ 3 [19]. Similarly, in third- or later-line treatment trials for LBCL, axi-cel had a higher rate of grade ≥ 3 neurotoxicity (28%) compared with tisa-cel (12%) and liso-cel (10%) [16,24,30].

Neurotoxicity has also been a notable concern in R/R indolent B-cell lymphoma, particularly with axi-cel. In the ZUMA-5 trial, neurologic adverse events occurred in 59% of patients (87 of 148), including 56% of those with FL and 71% with MZL [21]. Grade 1-2 events were observed in 40% of patients, while 19% experienced grade 3-4 events. The median onset was 7 days, with a median duration of 14 days for FL and 10 days for MZL. Conversely, liso-cel showed a much lower neurotoxicity profile in the phase 2 TRANSCEND trial, with any-grade events reported in 15% of patients, predominantly grade 1 (12%) and grade 3 in only three patients. The median onset was 8.5 days, and median duration was 3.5 days [33]. Similarly, the ELARA trial evaluating tisa-cel in FL reported any-grade neurologic events in 37.1% of patients, with ICANS occurring in only 4.1% within 8 weeks post-infusion [27]. Three patients developed grade ≥ 3 events, including one case of grade 4 ICANS. The median time to onset was 9 days, with resolution typically within 2 days. In R/R MCL, the pivotal phase 2 ZUMA-2 study found that brexu-cel caused neurotoxicity in 62% of patients, with grade 3-4 events in 31% [37]. By comparison, liso-cel in R/R MCL was associated with grade 3-4 neurologic events in only 9% of patients [36]. Overall, neurotoxicity is more frequent and severe with axi-cel and brexu-cel, particularly in R/R LBCL and MCL. Axi-cel consistently shows higher rates of neurological complications across different lymphoma subtypes, whereas liso-cel exhibits the lowest incidence and severity. Moreover, these events tend to be more common and serious in aggressive lymphomas such as LBCL and MCL compared to indolent B-cell lymphomas.

Neurologic adverse events associated with BsAbs are generally infrequent and milder compared to those seen with CAR T-cell therapy. Notably, BsAb-related neurotoxicity has not been linked to circulating BsAb molecules, activated T cells, or proinflammatory cytokines in the cerebrospinal fluid. While brain cells express CD19—the target of CAR T-cell therapies—they do not express CD20, the target of CD3×CD20 BsAbs [82]. This likely contributes to the lower incidence and severity of BsAb-induced neurotoxicity [80]. Reported neurologic events in BsAb trials have mostly consisted of mild symptoms such as headache and dizziness, with ICANS-like presentations occurring in only 1%-8% of cases. In third- or later-line treatment of LBCL, the phase 2 EPCORE NHL-1 trial reported ICANS in 6.4% of patients receiving epcoritamab (n=157), including grade 1 (4%), grade 2 (1%), and one grade 5 event (1%) [53]. Glofitamab was associated with ICANS in 9% of patients (n=154), with 6% experiencing grade 1-2 events and 3% having grade ≥ 3 events [63]. Mosunetuzumab induced ICANS in two patients (1.1% each), presenting as grade 1 confusion and attention disturbance; no grade ≥ 3 neurologic events were reported [51]. In a trial of odronextamab for R/R LBCL, no ICANS events were observed among 141 patients [78]. In patients with R/R FL, mosunetuzumab was associated with neurologic events in 5% of cases, all of which were grade 1-2 [48]. Epcoritamab induced ICANS in 6% of 128 R/R FL patients, again all grade 1-2 [54]. Odronextamab had a single case of grade 2 ICANS reported with the 0.7/4/20 mg step-up dosing schedule [59]. In R/R MCL, glofitamab given after pretreatment with two different doses of obinutuzumab (n=37) resulted in ICANS in 13.5% of patients, all grade 1-2 [79].

Infections associated with T-cell–mediated therapies for B-cell lymphomas span a wide clinical spectrum, from mild to potentially life-threatening. These include bacterial, fungal (notably Pneumocystis jirovecii), and viral pathogens such as cytomegalovirus, hepatitis viruses, respiratory viruses, and others. Heavily pretreated patients—often receiving CAR T-cell therapy as third-line or later treatment—are particularly susceptible to opportunistic infections. Clinical trials have consistently reported a notable incidence of severe (grade 3 or 4) infectious complications in this setting (Table 2). In the ZUMA-1 trial, 30 patients (28%) developed grade ≥ 3 infections, primarily lung infections and bacteremia, during the extended follow-up period [15]. Similarly, the JULIET trial documented grade 3 or 4 infections in 20% of patients, with 6% experiencing infections concurrently with CRS [23]. In the TRANSCEND NHL-001 study, 33 patients (12%) experienced grade ≥ 3 infections, including severe bacterial infections (4%), fungal infections (1%), and viral infections (1%) [29]. The use of CAR T-cell therapy in the second-line setting has also been associated with notable infection risks. In the ZUMA-7 trial, 17% of patients experienced grade ≥3 infections [20], while the TRANSFORM trial reported a similar incidence at 15% [32]. In the PILOT trial, grade 3 or higher infections occurred in 7% of patients (four individuals) [77].

The use of CAR T-cell therapy in indolent B-cell lymphomas warrants careful consideration, given the need to balance therapeutic benefits against potential risks, particularly infections. Among the available products, tisa-cel demonstrated a favorable safety profile in the ELARA trial, with only 5.2% of patients experiencing grade ≥ 3 infections within 8 weeks post-infusion [27]. In contrast, axi-cel in the ZUMA-5 trial was associated with a higher rate of severe infections, reported in 18% of patients (26 individuals), including pneumonia in 10 patients (7%) [21]. Similarly, liso-cel in the TRANSCEND FL trial showed grade ≥ 3 infections in 10.3% of patients [33].

In R/R MCL, infection remains a significant concern during CAR T-cell therapy. The ZUMA-2 trial reported a high incidence of grade ≥ 3 infections in 32% of patients, with pneumonia being the most frequent (9%) [37]. Notably, two grade 5 infection-related adverse events occurred—organizing pneumonia and Staphylococcus bacteremia—both associated with conditioning chemotherapy. In the MCL cohort of the TRANSCEND NHL-001 trial, grade ≥ 3 infections were observed in 13 patients (15%), including one death due to coronavirus disease 2019 (COVID-19) pneumonia and another due to cryptococcal meningitis [36].

Given the need for continuous administration, BsAbs also carry a significant risk of infectious complications in patients with R/R LBCL, with infection rates comparable to those seen with CAR T-cell therapies (Table 2). In clinical trials, grade ≥ 3 infections occurred in 14.6% of patients treated with epcoritamab, out of an overall infection rate of 45.2% [53]; in 15% of patients receiving glofitamab, among 59 patients with infections (38%) [63]; and in 12.5% of those treated with mosunetuzumab [51]. In contrast, odronextamab was associated with a notably higher rate of severe infections. In the ELM-2 trial, 37% of patients (52 individuals) experienced grade ≥ 3 infections, including 16 fatal (grade 5) cases (11%). COVID-19 infections were reported in 23 patients (16%), with six resulting in death (4%) [78]. In R/R FL, mosunetuzumab was associated with grade ≥ 3 infections in 13 patients (14%), with common etiologies including urinary tract infections, COVID-19 pneumonia, and Epstein-Barr virus viremia [48]. Notably, no infection-related grade 5 (fatal) adverse events were reported. In the EPCORE NHL-1 trial involving patients with FL, infections occurred in 17 patients (13%) treated with epcoritamab, with six fatal cases (5%) due to COVID [54]. For R/R MCL, glofitamab treatment was associated with four deaths attributed to COVID-19–related pneumonia [79].