Challenges and solutions to superior chimeric antigen receptor-T design and deployment for B-cell lymphomas

Summary

Chimeric antigen receptor-T (CAR-T) therapies represent a major breakthrough in cancer medicine, given the ex vivo-based technology that harnesses the power of one's own immune system. These therapeutics have demonstrated remarkable success for relapsed/refractory B-cell lymphomas. Although more than a decade has passed since the initial introduction of CAR-T therapeutics for patients with leukaemia and lymphoma, there is still significant debate as to where CAR-T therapeutics fit into the management paradigm, as consensus guidelines are limited. Competing interventions deployed in subsequent lines of therapy for aggressive lymphoma include novel targeted agents, bispecific antibodies, and time-honoured stem cell transplant. In this focused review, we discuss the major obstacles to advancing the therapeutic reach for CAR-T products in early lines of therapy. Such barriers include antigen escape, “cold” tumour microenvironments, host inflammation and CAR-T cell exhaustion. We highlight solutions including point-of-care CAR-T manufacturing and early T lymphopheresis. We review the evidence basis for early CAR-T deployment for B-cell lymphomas in light of the recent Food and Drug Administration (FDA) approval of three first-in-class anti-CD3/CD20 bispecific antibodies—mosunetuzumab, epcoritamab and glofitamab. We propose practical recommendations for 2024.

INTRODUCTION

Despite advancements in small molecule antagonists, immunomodulators and antibody therapies, many patients with aggressive lymphomas often still face relapsed or refractory (R/R) disease. The concept of adoptive T cell therapeutics was introduced more than three decades ago, when the idea of chimeric antigen receptor-T (CAR-T) cells came about in 1989.¹ Only recently have these efforts come to fruition at the bedside on a commercial scale. CAR-T cell therapy has revolutionized how patients with lymphoma are treated by allowing patients to receive a la carte care partly tailored to their cell surface proteome.² CAR-T intervention is sometimes considered as a personalized transplant, given the manufacturing schema which harnesses autologous immune effector mechanisms to target one's own cancer.

The anatomy of a CAR-T construct includes five major components—an extracellular antigen-binding domain, a hinge domain, a transmembrane domain, a co-stimulatory domain, and an intracellular signalling domain.² The antigen-binding domain, known as the single chain variable fragment (scFv), is located extracellularly. For commercially approved CARs, this scFv recognizes CD19 on malignant B cells and B-cell maturation antigen on malignant plasma cells. The hinge domain is important for the flexibility of the antigen-binding domain and tethers it to the transmembrane domain. The transmembrane domain allows for the downstream signalling cascade to initiate upon binding of the antigen. The co-stimulatory domain consists of either CD28 or CD137 (4-1BB) and functions to enhance the robust attack on the target cell. The intracellular domain usually consists of CD3ζ (the intracellular portion of the CD3 complex), which mediates effector T cell function.^2,3

Historically, the initial engineering efforts for CARs involved use of only the extracellular scFv and the intracellular CD3ζ activation domain—these comprised first-generation CARs. These first-generation CARs would depend solely on scFv for antigen binding and on CD3ζ activation domain for the effector response.^2,3 Further optimization of CAR design led to second-generation constructs, which included the components of a first-generation CAR and a co-stimulatory domain such as CD28 and 4-1BB. The current Food and Drug Administration (FDA)-approved CAR-T therapies are mostly second-generation CARs.^2,3 In an effort to enhance immune effector function, third-generation CARs were eventually developed, and these have at least two co-stimulatory domains. Fourth-generation CARs contain cytokine inducers and involve NFAT activation. Finally, fifth-generation CARs contain a cytokine receptor within the construct, allowing for JAK–STAT activation.³

Despite the success of CAR-T therapies, grade 3 and 4 toxicities are common with CAR-T products—these iatrogenic complications may sometimes lead to suboptimal outcomes. Cytokine release syndrome (CRS) and neurotoxicity are the most well-known toxicities of these therapies. Although numerous advances have been made in CAR-T design to help limit these toxicities, such complications merit additional work so that CAR-T therapeutics can be delivered as safely and effectively as possible.

Search strategy and selection criteria

References for this review were identified through searches of PubMed with the search terms “CAR-T therapies”, “B cell lymphoma”, “cell-based therapies” and “chimeric antigen receptor T” from 2005 to 2023. Articles were also identified through searches of the authors' own files. Only papers published in English were reviewed. The final reference list was generated on the basis of originality and relevance to the broad scope of this review.

CURRENT FDA APPROVALS OF CAR-T THERAPIES ACROSS NOVEL INDICATIONS IN LYMPHOMA

Axicabtagene ciloleucel

The first-in-class CAR-T therapy for aggressive lymphomas was axicabtagene ciloleucel (axi-cel) (formerly known as KTE-C19).^4,5 Axi-cel is approved for adults with R/R large B-cell lymphoma (LBCL) including diffuse large B-cell lymphoma (DLBCL), primary mediastinal large B-cell lymphoma (PMLBCL), and high-grade B-cell lymphoma (HGBCL). The objective response rate (ORR) from ZUMA-1 data was 82% and the complete response (CR) rate was 54%. Longer-term follow-up showed that the ORR was 83% and the CR rate was 58%.⁵ Axi-cel was then approved for R/R follicular lymphoma after two or more lines of systemic therapy based on ZUMA-5 data, which showed ORR of 92% and CR rate of 74%.⁶ It then gained approval in the second line setting for LBCL based on ZUMA-7 data, which showed superior median event-free survival (EFS) for axi-cel compared to investigator's choice standard-of-care (8.3 months vs. 2 months) and improved 2-year EFS (41% vs. 16%) (Table 1).⁷

TABLE 1.

Landmark clinical trials for currently available CAR-T therapies for B-cell lymphoma.

Clinical trials	Product	Phase of study	Disease	Study setting	FDA-approved cell dose	Response rate (ORR or CR)	Reference
ZUMA-1	Axi-cel	II	DLBCL	49/77 patients had ≥3 prior lines of therapy 23/77 had history of primary refractory disease 39/77 had history of resistance to two consecutive lines	2 × 106 viable CAR-T cells/kg	ORR: 82% CRR: 54%	Neelapu et al.4,8
ZUMA-5	Axi-cel	II	FL	Third line	2 × 106 viable CAR-T cells/kg	ORR: 92% CRR: 77%	Jacobson et al.6
ZUMA-7	Axi-cel	III	DLBCL	Second line	2 × 106 viable CAR-T cells/kg	CRR: 65%	Locke et al.7
ZUMA-12	Axi-cel	II	DLBCL	First line	(Not approved in the first-line setting)	ORR: 89% CRR: 78%	Neelapu et al.8
JULIET	Tisa-cel	II	DLBCL	Range from 1 to 6 previous treatments	0.6–6 × 108 viable CAR-T cells/kg	ORR: 52% CR: 40% PFS: estimated 83% at 12 months for those who had complete or partial response at 3 months OS: 12 months	Schuster et al.10
ELIANA	Tisa-cel	II	B-cell ALL	Fourth line (median of three prior therapies)	0.2–5 × 106 viable CAR-T cells/kg for patients weighing 50 kg and under 0.1–2.5 × 108 viable CAR-T cells/kg for patients weighing above 50 kg	ORR: 90% at 6 months, 76% at 12 months	Maude et al.11
ELARA	Tisa-cel	II	FL	Fifth line (median of four previous treatments)	0.6–6 × 108 viable CAR-T cells/kg	ORR: 86.2% CRR: 69.1%	Fowler et al.12
ZUMA-2	Brexu-cel	II	MCL	Sixth line (up to five previous treatments)	2 × 106 viable CAR-T cells/kg	Objective response rate: 85% CRR: 59%	Wang et al.13
ZUMA-3	Brexu-cel	II	B-cell ALL	Relapsed/refractory	1 × 106 viable CAR-T cells/kg or 1 × 108 max dose	Complete remission rate: 56%	Shah et al.14
TRANSCEND NHL 001	Liso-cel	I	DLBCL, FL Grade 3B	Fourth line (median of three previous treatments)	50–110 × 106 viable CAR-T cells if used in the third line setting and beyond	Objective response rate: 73% CRR: 53% Median PFS was 6.8 months	Abramson et al.15
TRANSFORM	Liso-cel	III	DLBCL	Second line in primary refractory or early (≤12 months) relapse	90–110 × 106 viable CAR-T cells if used in the second line setting	CRR: 74% Median PFS: not reached at 17.5 months median follow up	Kamdar et al.16; Abramson et al.17

Abbreviations: ALL, acute lymphoblastic leukaemia; axi-cel, axicabtagene ciloleucel; brexu-cel, brexucabtagene autoleucel; CAR-T, chimeric antigen receptor-T; CR, complete response; CRR, complete response rate; DLBCL, diffuse large B-cell lymphoma; FDA, Food and Drug Administration; FL, follicular lymphoma; liso-cel, lisocabtagene maraleucel; MCL, mantle cell lymphoma; MM, multiple myeloma; ORR, overall response rate; PFS, progression-free survival; tisa-cel, tisagenlecleucel.

Tisagenlecleucel

Tisagenlecleucel (tisa-cel) is approved for patients up to age 25 with R/R B-cell acute lymphoblastic leukaemia (B-ALL), adults with R/R DLBCL after two or more lines of therapy, and adults with R/R follicular lymphoma after two or more lines of therapy.^11,18 Tisa-cel for B-ALL was first studied in the global single-cohort phase II ELIANA study.¹¹ After at least 3 months following a single dose of tisa-cel, the overall remission rate was 81%. The 6-month relapse-free survival rate was 80%. In the international phase II JULIET trial, tisa-cel was studied in R/R DLBCL.¹⁰ Among those who received an infusion, the median overall survival (OS) was 12 months. Finally, tisa-cel was recently studied in adult patients with R/R follicular lymphoma in the multinational phase II ELARA trial.¹² At approximately 9.9 months follow-up, the CR rate was 69% and the ORR was 86%. At 12 months, the PFS rate was 67%.

Brexucabtagene autoleucel

Brexucabtagene autoleucel (brexu-cel) (formerly known as KTE-X19) is approved for adults with R/R mantle cell lymphoma and R/R B-ALL based on results of the ZUMA-2 and ZUMA-3 trials, respectively.^13,14 In the ZUMA-2 study, the ORR and CR rates were 85% and 59%, respectively.¹³ In the ZUMA-3 study, the CR rate was 56%, and the median time to reach CR or CR with incomplete haematological recovery was 1.1 months.¹⁴

Lisocabtagene maraleucel

Lisocabtagene maraleucel (liso-cel) is FDA-approved for adult patients with R/R LBCL in second line setting and beyond, including R/R DLBCL, PMLBCL and HGBCL, as well as for follicular lymphoma stage 3B.^15-17 In the TRANSCEND NHL 001 trial, the ORR was 73%, and CR rate was 53%.¹⁵ The median OS was 21.1 months, and the median PFS was 6.8 months.¹⁵ In the Phase III TRANSFORM trial, liso-cel showed improved median PFS compared to standard 2nd line therapy (10.1 months vs. 2.3 months).¹⁶ Very recent primary analysis of the TRANSFORM trial showed that the median EFS was not reached at 17.5 months of median follow up, compared to 2.5 months for standard second-line therapy.¹⁷ The CR rate was also improved (74% vs. 43%).¹⁷

Real-world data

Since 2017, four FDA-approved CAR-T cell therapies have been used in clinical practice for a variety of lymphoma subtypes. The use of CAR-T therapies in the real world (RW) setting allows for comparison with clinical trials. Overall, CAR-T cell therapy has been beneficial to patients who did not participate in the registration studies due to trial ineligibility. In a retrospective cohort study analysing the efficacy of CAR-T therapy in patients with DLBCL, there were lower rates of adverse events, including CRS and anaemia, compared to the respective clinical trial.¹⁹ At 12 months, the survival rate for RW patients was 59%, while the rate from clinical trials for tisa-cel, axi-cel and liso-cel were 49%, 59% and 58%, respectively.¹⁹ Similarly, 168 RW patients with mantle cell lymphoma who received brexu-cel experienced similar efficacy and toxicity when compared to patients enrolled in the ZUMA-2 trial.²⁰ The 6-month PFS was 69%, and 12-month PFS was 59%.²⁰ RW patients could receive bridging therapy ranging from venetoclax to radiation whereas patients in the ZUMA-2 study could only receive BTKi or corticosteroids. The difference in bridging therapy could account for the different CR rates in RW patients compared to patients from registration trials.

The success of CAR-T therapy in the RW setting has been reported internationally. In 21 sites across Germany, 356 RW patients received axi-cel or tisa-cel to treat R/R LBCL.²¹ While the OR, CR, and OS rates between the RW patients and patients enrolled in the ZUMA-1 and JULIET trials were similar, the PFS for RW patients was lower at 35% and 24%, for axi-cel and tisa-cel, respectively. The RW patients had lower performance status with a greater need for bridging therapy, which could explain why the PFS was lower. The encouraging RW data points to the feasibility of CAR-T therapy outside of registration studies.

CHALLENGES AND PROPOSED SOLUTIONS TO SUPERIOR CAR-T DESIGN AND EARLY DEPLOYMENT

Despite the widespread success of CAR-T therapy in treating patients with lymphomas, leukaemias and multiple myeloma, more work remains to be done to address the numerous challenges that have come to the forefront of clinical care. These challenges include antigen escape, tumour microenvironmental barriers, host factors, T cell factors, toxicity, high costs, manufacturing time concerns, as well as limited knowledge about the best clinical setting (i.e. line of therapy) in which CAR-T cells should be deployed. We herein summarize the current knowledge about these problems and highlight rational solutions.

The first major challenge is antigen escape, which inevitably leads to limited durability of response to CAR-T and refractoriness (Figure 1). Patients with recurrent disease after treatment harbour cells that have diminished levels of CD19, thus excluding them benefiting further from CAR-T therapies.^22,23 Antigen loss can occur through disruptive mutations in the open reading frame of CD19 as well as alternative splicing. Antigen masking also contributes to resistance. One solution to this problem is to employ dual- or multiantigen targeted CAR-T cells, as these CARs would recognize another antigen separate from CD19.²⁴ The efficacy of multiantigen targeted CAR-T cells may not depend on presence of CD19. In a phase I clinical trial led by Stanford, a bispecific CAR-T therapy (targeting CD19 and CD22 in the setting of R/R B-ALL and LBCL) was assessed.²⁵ Response rate was 100% for B-ALL, with 88% measurable residual disease (MRD)-negative CR rate. Response rate was 62% for LBCL, with 29% CR rate. Adapter CAR-T technology using tumour-specific adapter molecule combinations may overcome the problem of antigen escape.²⁴ These adapter CARs may target alternative antigens like CD20, CD79B or CD22. The problem of antigen escape calls for expansion of antigens recognized by CAR-T cells.

FIGURE 1.

Barriers and solutions to optimal deployment of CAR-T therapeutics. Major barriers include host factors, T cell factors and the tumour microenvironment. Such barriers must be overcome to facilitate earlier deployment of CAR-T therapeutics. Ag, antigen; CAR-T, chimeric antigen receptor-T; MDSC, myeloid-derived suppressor cell; TME, tumour microenvironment; T_reg, regulatory T cell.

Another major barrier is the tumour microenvironment (TME), which may hinder CAR-T cell efficacy in certain lymphomas. For example, the TME might prevent CAR-T cells from mobilizing to the site through dense extracellular matrix and vasculature that promotes hypoxia.²⁶ A hypoxic environment promotes immunosuppressive cell infiltration and downregulation of adhesion molecules that help CAR-T cells mobilize to the site. Expansion of CAR-T cells under hypoxia may promote memory T cells.²⁷ Endogenous immune checkpoints can lead to unwanted CAR-T cell inhibition. Depletion of regulatory T cells (T_regs) or myeloid-derived suppressor cells within the TME may facilitate CAR-T efficacy.^28,29 Efforts at combating the “cold” TME may need to focus on extracellular matrix remodelling to enhance CAR-T cell infiltration.³⁰ In addition, the immunosuppressive cytokine milieu within the TME can hamper CAR-T cell function. Inhibitory ligand expression in the TME can also hinder CAR-T cells.

Host factors such as inflammation also pose a challenge to successful CAR-T administration. In inflammatory conditions, T cells can become exhausted due to the chronic stimulation by antigens.³¹ Thus, patients with active inflammation who are undergoing CAR-T infusion are less likely to benefit from their therapy compared to those without active inflammation. Immunomodulatory issues related to host inflammation can also arise: short in vivo persistence may be related to presence of T_regs. These CAR-T_reg cells can be present shortly after infusion of the CAR-T product and can lead to progressive disease.³² To overcome some of these obstacles, efforts may need to focus on enhancement of T cell activation. The interaction between programmed cell death (PD-1) receptor and its ligand, programmed cell death ligand (PD-L1), can be therapeutically blocked. Indeed, CAR-T exhaustion was diminished when a dominant negative PD-1 receptor was expressed.³³ Another method is to perturb the CAR-T and antigen interaction.³⁴ This would require engineering the CAR-T to have optimal antigen interaction such that it still has strong therapeutic effects but disrupts the constant stimulation by antigen interaction.

T cell factors may also pose a hurdle to clinical rollout of more effective cell-based products. Commercial CAR-T therapies for lymphoma currently target only CD19, but there is high theoretical potential to expand the target antigen repertoire. Generally, the scFv is responsible for antigen specificity and thus partly influences CAR-T cell activation. This warrants the need for the expansion of antigens that CAR-T cells can recognize. CD20 and CD22 have been studied as alternative targets as these are lineage-specific B-cell markers. CD22-directed CAR-T therapy has shown promising results in the single-agent setting for B-ALL and when combined with CD19-directed CAR-T cells.³⁵

Another factor that may pose challenges is high toxicity of CAR-T cells. CAR-T therapeutics notably lead to cytopenias, CRS and neurotoxicity. There are fewer than a handful of therapies to combat such adverse effects. Some solutions to this issue of high toxicity include use of termination mechanisms or transiently active mRNAs encoding CARs, creation of rapidly switchable platforms (employment of an “on/off” switch), engineering of biodegradable CARs and ablation of T cells post-infusion.^36-39

A major obstacle to the widespread adoption of CAR-T therapy use is the cost and time associated with manufacturing. In the United States, the average expected cost is approximately $500 000.⁴⁰ To address this problem, institutions can manufacture their own CAR-T products, known as point-of-care manufacturing. While point-of-care manufacturing both streamlines the production process and allows for greater accessibility, it requires robust and standardized quality control measures. One solution is to use a standardized, semiautomated platform. The ARI-0001 CAR-T production system was the first to show the feasibility of CAR-T manufacturing using a closed semi-automatic bioreactor platform.⁴¹ In an academic Phase I clinical trial, 27 patients with ALL, CLL and NHL underwent T cell harvest to generate anti-CD19 CAR T cells.⁴¹ Of all 28 products (one patient had apheresis twice due to production failure), the average frequency of CAR+ cells was 30.6%. There was no difference in efficiency of transduction between T cells collected from patients with different diseases. Moreover, all manufactured CAR-T cells could eliminate CD19+ target cells to <70%.⁴¹

In addition to point-of-care manufacturing to provide CAR-T therapies, providers must also consider the timing of T cell harvest. Currently, only axi-cel and liso-cel are the FDA-approved CAR-T therapies used in second-line treatment of LBCL. As patients undergo multiple lines of therapy, both the quantity and quality of the harvested T cells may diminish. An effective approach to overcome this is earlier lymphopheresis to harvest healthier T cells. In 2023, a prospective study was conducted investigating the effects of earlier lymphopheresis in patients with R/R DLBCL.⁴² Twenty-two patients with R/R DLBCL had lymphopheresis after failure of first-line therapy or relapse after one line of treatment, while 25 patients with R/R DLBCL received lymphopheresis after a median of three lines of treatment.⁴² Patients who had an earlier harvest time had an increased percentage of naive T cells and lower T cell exhaustion compared to patients with a later harvest time. Moreover, the ORR for patients with an earlier harvest was 60%, whereas the ORR for patients with a later harvest was 35%. In the Bio-CAR-T BS study, patients who received early pre-emptive lymphocyte apheresis had a significantly different T cell population compared to those who underwent standard lymphocyte apheresis. Specifically, early T cell collection resulted in a higher CD4+/CD8+ ratio, higher CD4+ naive T cells, lower CD4+ effector memory T cells, and lower CD8+ terminally differentiated cell when compared to T cells collected in the standard group.⁴³

Finally, a considerable challenge is that we do not have long-term follow-up data to know which clinical setting (i.e. line of therapy) is optimal for delivering CAR-T cells. This knowledge will have major impacts on sequencing of therapies for patients. Currently, CAR-T cell therapy is approved in the fifth-line setting for multiple myeloma, and until only very recently, CAR-T therapies had been available only in the third-line setting and beyond for aggressive lymphomas. There is some debate about the value of offering CAR-T therapy in the earlier lines of care. Last year's BELINDA trial, for example, showed that tisa-cel was not superior to standard-of-care in the second-line setting for aggressive lymphomas.⁴⁴ However, last year's ZUMA-7 showed benefit for axi-cel in the second-line setting, leading to FDA approval.⁷ Liso-cel is also approved in the second-line setting based on the TRANSFORM trial and may be superior to traditional salvage chemotherapy options in select patients.¹⁶

Very recent results from the ZUMA-12 study may change this paradigm of reserving CAR-T therapy for later lines of therapy. ZUMA-12 showed that axi-cel is a successful first-line therapy for patients with high-risk LBCL, with ORR of 89% and CR rate of 78%.⁸ The success of axi-cel as first-line therapy calls for further investigation into its benefits compared to time-honoured, first-line anthracycline-based chemotherapy options. One barrier to moving CAR-T therapy to first-line treatment is its high rate of adverse effects and perhaps high risk for tumour lysis in the absence of initial debulking. In the ZUMA-12 study, 100% of treated patients unfortunately had an adverse event, and 85% of the total treated patients experienced grade ≥3 adverse events.⁸ All the treated patients also had CRS of any grade, with 8% of the total treated patients having grade ≥3 CRS. Of all treated patients, 73% had neurological events of any grade. The high death rate (15%) in the first-line setting, however, may not justify its use in select patients. It will be important to know the rates and durability of MRD-negative remissions with early use of CAR-T therapy.

A barrier to early deployment of CAR-T therapies is the recent introduction of novel competing therapies including three bispecific T cell engagers (BiTEs). Bispecific antibodies have two binding sites that target two different antigens or two different epitopes on the same antigen. Compared with CAR-T products which are personalized medicine and hence associated with an extensive manufacturing process, bispecific antibodies are ready off-the-shelf. Currently, three agents are FDA-approved for lymphoma in the R/R setting; these include mosunetuzumab, epcoritamab and glofitamab. These agents target both CD3 on T cells and CD20 on B cells. Mosunetuzumab is an IgG1-based bispecific antibody that was approved in December 2022 for R/R follicular lymphoma.⁴⁵ Ninety patients with follicular lymphoma were enrolled and received a median of eight cycles. The ORR was 80% and the CR rate was 60%. Epcoritamab was approved in May 2023 for R/R DLBCL and high-grade B-cell lymphoma based on the results of the EPCORE NHL-1 trial.⁴⁶ A total of 157 patients with R/R CD20+ DLBCL or other aggressive non-Hodgkin lymphoma were enrolled and received a median of five cycles. The ORR was 63%, and the CR rate was 39%. Glofitamab was approved in June 2023 for R/R DLBCL after two or more lines of therapy. Glofitamab is unique due to its bivalency for CD20 and monovalency for CD3.⁴⁷ A total of 155 patients were enrolled, and CR rate was 39%, with 12-month PFS was 37%. One additional investigational agent is odronextamab for which additional clinical trial data is pending.⁴⁸ Head-to-head comparisons of CAR-T therapies and bispecific antibodies will be valuable in the R/R setting, as each modality has its unique benefits and risks, as discussed below.

SUMMARY OF BEST AVAILABLE EVIDENCE

The therapeutic landscape of cell-based therapies has evolved significantly since the first demonstration of its efficacy more than a decade ago in children. As of 2023, we now have a robust evidence basis for high-value care involving CAR-T therapy in the second-line setting for aggressive lymphomas, and we may soon see approval of CAR-T therapy in the front-line setting. Future efforts should focus on optimal CAR design such that we can enhance efficacy (via improved in vivo durability and lower risk for antigen escape), reduce toxicity, and optimally position such therapies in favour of durable remissions. Superior CAR-T design strategies will likely need to integrate functional genomics with bioengineering including novel armoring systems, and the efficacy of such novel approaches can likely be enhanced by ensuring optimal immunomodulatory conditions in patients at the time of CAR-T delivery.

With respect to the setting of deployment, the introduction of CAR-T therapy in the first- or second-line settings is highly appealing, especially for patients whose goal of therapy is to proceed with curative intent. Very recent 5-year follow-up data for axi-cel, reported by Neelapu et al., showed 5-year OS of 42.6% and alluded to the curative potential for axi-cel.⁹ Ongoing responses at the 5-year mark from ZUMA-1 have been associated with expansion of CAR-T cells.⁹ This attests to the sustainability of CAR-T therapies. The model of using CAR-T therapy as a bridge to stem cell transplant may be replaced by CAR-T as destination therapy if long-term data lends support to this. There is strong data to favour introduction of appropriate CAR-T therapies in earlier lines of care, including the second-line setting, given the high response rates and curative potential for this modality. Of note, patients may need sufficient debulking with cytotoxic chemotherapy prior to CAR-T infusion, given the risk for CRS and neurotoxicity if disease burden is high. In addition to the autologous products on the market, allogeneic CAR-T cells (off-the-shelf products) may provide therapeutic benefit. The use of banked allogeneic CAR-T cells bypasses the need for T cell harvest from a patient who may have insufficient cells to provide, and it also expedites the process by bypassing manufacturing time of an autologous product. However, histocompatibility barriers will need to be addressed with allogeneic CAR-T cells, as there is a possibility of graft-versus-host disease and graft rejection.

The recent FDA approval of the three first-in-class bispecific antibodies for B-cell lymphoma offers another valuable option in the R/R setting. Bispecific antibodies may be used in select patients who are physiologically unfit to tolerate an intense one-time dose of curative intent CAR-T cells or in patients with significant logistical barriers to receiving CAR-T therapy. A downside of bispecific antibodies is that response rates are generally lower than CAR-T therapies, and no mature clinical data has alluded to curative potential. Therapy with bispecific antibodies may be indefinite, rather than time-limited like CAR-T therapy. Furthermore, the issue of T cell exhaustion in BiTE-exposed patients before CAR-T therapy needs to be explored, as there is some data in multiple myeloma that BiTE exposure can impair CAR-T outcomes. This idea may influence selection and sequencing of CAR-T therapy in the coming years, and such considerations merit large-scale clinical trial efforts.

Despite the recent advances in translational science for adoptive T cell therapies for B-cell lymphomas, major issues in this area that merit further investigation include multiantigen targeting, novel B-cell target validation, allogeneic CAR construction, manufacturing innovation, and incorporation of gene therapy with cell-based therapy, such as design of base-edited CARs.⁴⁹ In the coming years, we anticipate that investment into such strategies will improve translational science and will help expand the therapeutic reach of CAR-T therapy, leading to superior outcomes for patients with R/R B cell lymphomas.⁵⁰