Dual-target CAR-T cell therapy: latest updates from the 2025 ASH annual meeting

Abstract

Dual-target chimeric antigen receptor T (CAR-T) cell therapy has emerged as a promising strategy to address the key limitations of single-target CAR-T approaches, including antigen heterogeneity and antigen-negative escape. Recent studies have reported rapid progress in the engineering of dual-target CAR-T, incorporating diverse receptor architectures, logic-gated designs, and novel manufacturing platforms. Preclinical data demonstrate improved disease control across B-cell malignancies, acute myeloid leukemia, and autoimmune disease models. Here, we summarize the latest advancements in the preclinical investigations and clinical trials of dual-target CAR-T cell therapy presented at the 2025 ASH annual meeting.

To the editor,

Chimeric antigen receptor T (CAR-T) cell therapy has demonstrated remarkable clinical efficacy in various hematological malignancies, especially in B-cell-derived cancers. However, with increasing clinical experience, some limitations of single-target CAR-T cell therapy have gradually emerged [1]. Dual-target CAR-T cell therapy improves tumor coverage and reduces antigen-negative escape in hematological malignancies by simultaneously recognizing two tumor-associated antigens to address antigen heterogeneity and disease complexity [2]. Here, we highlight the latest progress in dual-target CAR-T therapy reported at the 2025 ASH annual meeting.

Engineering strategies for dual-target CAR-T cells

Recent studies indicate continued evolution in the engineering of dual-target CAR-T cells (Table 1). Conventional tandem CAR or dual-CAR architectures remain widely used, with ongoing optimization of single-chain variable fragment (scFv) order, linker design, and costimulatory domains to improve receptor expression and function [3]. Synthetic Notch receptor (SynNotch)-based AND-gate systems have been applied in acute myeloid leukemia (AML) models to restrict activation to dual-antigen-positive cells, thereby enhancing specificity and potentially reducing hematopoietic toxicity [4]. Other strategies, including dual-CAR co-expression with gene editing and intein-based modular receptor assembly, further aim to balance efficacy, safety, and payload feasibility [5–7].

Table 1.

Engineering strategies of dual-target CAR-T cells

Disease context	Target antigens	CAR architecture	Study stage
B-cell lymphoma	CD79b / CD19	Dual-target CAR	Preclinical
B-cell malignancies	CD19 / CD22	Dual-target CAR + metabolic modulation	Clinical translational
AML	CD33 / CD123	Dual-target CAR (multiple formats)	Preclinical
AML	CD70 / CD33	Dual-target CAR + secreted BiTE	Preclinical
AML	CD123 / CLL-1	SynNotch AND-gate CAR	Preclinical
AML	Tim-3 / CD70	Dual-CAR co-expression	Preclinical
Autoimmune diseases	CD19 / BCMA	Tandem/compound CAR	Preclinical
Lymphoid malignancies	Representative targets (CD19 / CD20)	Modular multi-receptor CAR/CCR	Preclinical

AML: acute myeloid leukemia; AND-gate: logical AND-gated receptor system; BCMA: B-Cell maturation antigen; BiTE: bispecific T cell engager; CAR: chimeric antigen receptor; CCR: chimeric costimulatory receptor; CLL-1: C-type lectin-like molecule-1; SynNotch: Synthetic Notch receptor; Tim-3: T cell immunoglobulin and mucin domain-containing protein 3

The CD33/CD123 AML study further showed how architecture can shape function. Co-transduction, bicistronic, tandem, and loop designs were active, but co-transduced cells retained stronger activity under low antigen density, suggesting no single architecture is universally optimal, and that design selection should be guided by disease context and target biology [8]. Beyond receptor design, metabolic optimization is emerging as an additional strategy. CD19/CD22 clinical samples revealed enhanced oxidative phosphorylation and memory-associated transcriptional profiles, informing solute carrier-armored “MetaboArm” CAR-T cells to improve persistence and antitumor activity [9].

Preclinical progress

In hematological malignancies, preclinical studies of dual-target CAR-T cells have focused mainly on B-cell malignancies, AML, and autoimmune disease models (Table 2). In B-cell lymphoma (BCL), CD79b/CD19 CAR-T cells showed stronger tumor clearance than single-target controls, especially in mixed antigen-expression models, supporting their value in overcoming intratumoral heterogeneity [3]. AML forces a trade-off between antigen coverage and hematopoietic safety. CD33/CD123 CAR-T cells showed activity across multiple engineering formats, with co-transduced cells performing better at low effector-to-target ratios [8]. Additional approaches combined a CD70-targeted CAR with secretion of a CD33×CD3 bispecific T-cell engager and gene-editing strategies to broaden coverage while attempting to reduce hematopoietic toxicity [5]. In parallel, a SynNotch CD123/CLL-1 AND-gate system restricted effector activation to dual-antigen-positive cells and achieved effective leukemia control in vivo [4], whereas Tim-3/CD70 dual-target CAR-T cells showed durable leukemia clearance in syngeneic models without sustained hematotoxicity [6]. Beyond malignancy, lipid nanoparticle (LNP)-mediated mRNA delivery of CD19/BCMA dual CAR-T cells provided a transient-expression strategy for autoimmune diseases, emphasizing safety control rather than permanent CAR expression [10]. Taken together, these studies suggest that the most useful dual-target format may be the one that balances specificity, rather than maximizes overall signaling.

Table 2.

Preclinical and clinical evidence of dual-target CAR-T cell therapies across disease settings

Disease	Target strategy	Model / study type	Safety observation	Clinical data	References
BCL	CD79b / CD19	In vitro and xenograft models	Not reported	NA	[3]
AML	CD33 / CD123	In vitro AML models	Not reported	NA	[8]
AML	CD70 / CD33	In vitro mixed-antigen models	Gene-editing strategy to reduce hematopoietic toxicity	NA	[5]
AML	CD123 / CLL-1	In vivo leukemia models	Reduced impact on normal hematopoietic cells	NA	[4]
AML	Tim-3 / CD70	Syngeneic mouse models (BM185, C1498)	No sustained hematotoxicity observed	NA	[6]
Autoimmune diseases	CD19 / BCMA	In vitro and animal models	Transient CAR expression to limit prolonged exposure	NA	[10]
BCL	CD19 / CD20	Phase I clinical trial (NCT05388695)	Grade ≥ 3 CRS 2%; no ≥ 3 ICANS	ORR 84.6%; CR 53.8%; 2-year PFS 69.2%; 2-year OS 76.2%.	[11]
BCL	CD19 / CD20	Ongoing clinical study (NCT05826535)	Not reported	Higher Cmax, AUC, delayed Tmax; enriched memory phenotype	[12]
BCL	CD19 / CD22	Phase II clinical trial (CAR2219; NCT06081478)	CRS 74% with no grade ≥ 3 CRS; ICANS 6% with one grade 3 event; no treatment-related death	ORR 100%; CR 64.5%; 12-month PFS 60.4%; 12-month OS 87.1%	[13]

AML: acute myeloid leukemia; AUC: area under the curve; BCL: B-cell lymphoma; BCMA: B-Cell maturation antigen; B-NHL: B-cell non-Hodgkin lymphoma; CAR: chimeric antigen receptor; CLL-1: C-type lectin-like molecule-1; CR: complete response; CRS: cytokine release syndrome; ICANS: immune effector cell-associated neurotoxicity syndrome; ORR: overall response rate; OS: overall survival; PFS: progression-free survival; Tim-3: T cell immunoglobulin and mucin domain-containing protein 3; T_max: time to maximal expansion

Clinical trials

Among reported studies, only a limited number of dual-target CAR-T programs have entered clinical evaluation with available efficacy data (Table 2). In a phase I trial (NCT05388695), CD19/CD20 tandem CAR-T cells (CAR2019[AS]) achieved an overall response rate (ORR) of 84.6% and a complete response (CR) rate of 53.8% in 52 evaluable patients with relapsed or refractory BCL; grade ≥ 3 cytokine release syndrome (CRS) occurred in 2%, with no grade ≥ 3 immune effector cell-associated neurotoxicity syndrome (ICANS) [11]. Another CD19/CD20 product (NCT05826535) showed higher peak expansion, greater exposure, delayed maximal expansion, and a more memory-like phenotype than approved CD19 products [12]. For CD19/CD22 targeting, CAR2219 achieved an ORR of 100% and a CR rate of 64.5% in relapsed or refractory large B-cell lymphoma (LBCL), with no grade ≥ 3 CRS, ICANS occurred in 6% of patients, including one grade 3 event [13]. Together, these data suggest meaningful activity for both CD19/CD20 and CD19/CD22 strategies, although cross-trial comparisons should remain cautious.

Translational and safety considerations

Several translational considerations remain. First, on-target off-tumor toxicity is highly dependent on antigen selection, particularly in AML [2, 4–6]. Second, dual-antigen engagement may alter the magnitude and duration of immune activation, thereby shaping CRS and ICANS profiles. However, these safety outcomes are influenced by the target pair, product format and disease burden [11–13]. Third, increased signaling complexity may predispose to tonic signaling, exhaustion, or imbalanced antigen sensitivity, potentially affecting persistence. Finally, greater construct complexity may also increase cost, variability, and production failure risk [2]. These factors highlight that the translational feasibility of dual-target CAR-T therapy relies on scalable, controllable designs that balance receptor architecture, antigen density, and safety, rather than on broader antigen coverage alone.

Limitations and future directions

Because many of the studies discussed here were presented in abstract form at the 2025 ASH Annual Meeting, current evidence remains preliminary, with limited methodological detail, short follow-up, and immature safety data. Nevertheless, these reports indicate that dual-target CAR-T therapy is moving toward refined construct design, improved safety control, and better manufacturing feasibility, but its optimal clinical role remains to be defined.

Abbreviations

AML

Acute myeloid leukemia

AND-gate

Logical AND-gated receptor system

AUC

Area under the curve

BCMA

B-cell maturation antigen

BCL

B-cell lymphoma

BiTE

Bispecific T cell engager

CAR

Chimeric antigen receptor

CAR-T

Chimeric antigen receptor T

CCR

Chimeric costimulatory receptor

CLL-1

C-type lectin-like molecule-1

CR

Complete response

CRS

Cytokine release syndrome

ICANS

Immune effector cell–associated neurotoxicity syndrome

LBCL

Large B-cell lymphoma

LNP

Lipid nanoparticle

NHL

Non-Hodgkin lymphoma

ORR

Overall response rate

OS

Overall survival

PFS

Progression-free survival

scFv

Single-chain variable fragment

SynNotch

Synthetic Notch receptor

Tim-3

T-cell immunoglobulin and mucin domain-containing protein 3

Author contributions

S.F. drafted this manuscript. S.F. and R.K. prepared tables. L.X. and X.Z. revised the manuscript. X.Z. provided direction and guidance throughout the preparation of the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by National Natural Science Foundation (No.82570247).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.