Main text
While single-antigen-targeting chimeric antigen receptor T cell therapy (CAR-T) against the B cell antigen CD19 has revolutionized the treatment landscape for relapsed/refractory non-Hodgkin lymphoma (NHL), a substantial proportion of patients still relapse following therapy, due to antigen-negative escape in around one-third of relapse cases.1 A variety of multi-antigen-targeting CARs for NHL have been tested in clinical trials, but the heterogeneity between dual-targeting product designs makes direct cross-trial comparisons difficult. Furthermore, it is still unclear what constitutes the optimal antigen combination and which is the most effective dual-targeting design approach.
In this issue of Molecular Therapy, Bachiller et al. report on the development and preclinical assessment of ARI0003,2 a novel dual-targeting CD19/B cell maturation antigen (BCMA) CAR-T cell therapy for NHL, manufactured using co-transduction and designed for enhanced therapeutic efficacy and to prevent CD19 antigen-negative escape. While BCMA is more classically associated with myeloma therapeutics, the authors indicate that BCMA is expressed on most lymphomas, making it a good CAR-T target.3 They show that their BCMA binder has significant antitumor effects on NHL cell lines with low BCMA expression.
The right dual-targeting design
Mono-targeted CAR-T is inherently limited by dependence on a single antigen, and dual-targeting CAR-T represents a logical evolution in the field. Several dual-targeting designs are outlined below and illustrated in Figure 1. Bicistronic vectors encode two unique CARs with two distinct signaling endodomains within a single construct, typically separated via self-cleaving peptide. Tandem CARs bind two antigens via a chain of two unique single-chain variable fragments (scFvs) connected by a linker and integrated into an otherwise conventional CAR architecture in the same manner as a single scFv binder. Loop CARs express an alternating sequence of two scFv light and heavy chains dimerized to form a loop structure.4 In contrast, co-transduction CARs (as per ARI0003) are the result of T cell incubation with a mix of two viral vectors, each encoding a separate CAR, and from which a mix of CAR-T populations (both mono- and dual-targeting) are generated. All of these designs have been previously tested and associated with some degree of clinical success in trials.5,6,7
Figure 1.
Different approaches for co-targeting of CD19 and BCMA in NHL using CAR-T therapy. Vector copy number (VCN) and surface CAR expression levels are represented (not in scale) with (+) and (↑) based on the findings of the authors during preclinical development of ARI0003 (Figure 3, E–H, Bachillier et al.2). Single transduced and co-transduced CAR-T cells have higher surface CAR expression compared with tandem and bicistronic CAR-T constructs, despite having lower VCN.
How does ARI0003 compare to other CD19/20/22 designs?
This article provides compelling preclinical evidence supporting the superiority of ARI0003 over mono-targeting CAR-T approaches. Briefly, ARI0003 demonstrated superior tumor control and overall survival in vivo, particularly in the low CD19 antigen density tumor setting. This is potentially clinically significant, as it suggests that dual-targeting CAR-T may retain efficacy even in the context of antigen heterogeneity/loss, a frequent challenge in the relapsed/refractory setting.
Furthermore, the authors show that ARI0003 delivered more effective antitumor activity than dual-targeting CD19 plus CD20/CD22 approaches.2 While it is possible that BCMA is simply an excellent target for NHL despite the low antigen density, it is also possible that the BCMA binder in this analysis is high sensitivity and/or more stable and thus more effectively eliminates antigen-low targets. Notably, the authors used unequal MOIs of the CD19/BCMA vectors to achieve matched surface CAR expression, as CD19CAR was consistently under-expressed compared with BCMACAR at equal MOIs. They hypothesized that this was due to competition for transcription/translation machinery within the cell.
Testing ARI0003 in different dual-targeting formats
To address the critical question of optimal dual-targeting CAR design, the authors tested these head to head with mono-targeted CARs in preclinical models using the same CD19/BCMA binder domains engineered into different formats. They then directly compared genomic integration and membrane trafficking/cell surface CAR expression between the different designs.
Notably, qPCR revealed that tandem and loop CAR designs had much lower cell surface expression than mono-CAR, with a 10-fold higher vector copy number (VCN) requirement to achieve equivalent expression levels. Similarly, bicistronic CARs had a 2-fold higher VCN requirement vs. mono-CAR.
They then tested whether higher VCN correlated with higher transcripts and, surprisingly, observed only a 1.5- to 3-fold increase in transcripts in tandem and loop CARs vs. mono-targeting CARs, potentially suggestive of transcriptional errors. Both ARI0003 and the bicistronic construct had transcripts comparable to those of mono-targeting CARs.
Comparing protein expression as measured by CAR mean fluorescence intensity (MFI) on the cell surface, dual CARs (tandem, loop, and bicistronic), and ARI0003 exhibited 60%–80% and 25% reduction in MFI, respectively, when compared with mono-targeting CARs. Taking this one step further to address whether protein trafficking to the cell surface was impaired, surface-to-total cell levels showed a 15%–20% reduction in trafficking of tandem and loop CARs to the cell surface when compared to other approaches.
Overall, these data suggest that the burden imposed on cells by dual-targeting CAR constructs can lead to reduced CAR expression, possibly from limitations in cellular transcription and translation machinery. However, it is unlikely that this fully explains the observed reductions in cell-surface protein expression seen here, hinting toward a broader set of challenges for the dual-targeting CAR-T field.
Summary
The field of CAR-T cell therapy is evolving rapidly, and dual-targeting strategies such as ARI0003 exemplify the innovative approaches that are needed to overcome resistance and relapse. A strength of ARI0003 lies in its versatility. The therapy is shown to be effective not only in antigen-rich environments but also in settings where antigen densities are low, a characteristic often associated with treatment failure.
Furthermore, the article by Bachiller et al. highlights the ability of ARI0003 to be manufactured under Good Manufacturing Practice conditions, which is critical for its translation into clinical practice.
While the preclinical results are promising, several challenges must be addressed to maximize the clinical impact of ARI0003. First, the exhaustion of dual-CAR-T cell populations over time, as observed in the study, underscores the need for further optimization to enhance their longevity. Strategies such as engineered modulation of T-cell signalling pathways and/or inclusion of alternative co-stimulatory endodomains are some of the many approaches that could be explored to sustain CAR-T cell activity.7
Second, the long-term efficacy and safety of dual-targeting CAR-T cells still needs to be validated in clinical settings. The ongoing phase I clinical trial of ARI0003 in NHL (study registered at ClinicalTrials.gov: CARTD-BG-01) will provide crucial insights into the performance of this therapy in patients with r/r NHL. These trials will also be essential in determining the optimal dosing and infusion protocols for ARI0003. By combining novel antigen targeting with an optimised dual-targeting engineering approach, ARI0003 heralds a new era in the treatment of r/r NHL.
Declaration of interests
The authors declare no competing interests.
References
- 1.de Oliveira Canedo G., Roddie C., Amrolia P.J. Dual targeting CAR-T cells for B-cell acute lymphoblastic leukaemia and B-cell non-Hodgkin lymphoma. Blood Adv. 2024 doi: 10.1182/bloodadvances.2024013586. [DOI] [PubMed] [Google Scholar]
- 2.Bachiller M., Barceló-Genestar N., Rodriguez-Garcia A., Alserawan L., Dobaño-López C., Giménez-Alejandre M., Castellsagué J., Colell S., Otero-Mateo M., Antoñana-Vildosola A., et al. ARI0003: Co-transduced CD19/BCMA dual-targeting CAR-T cells for the treatment of non-Hodgkin lymphoma. Mol. Ther. 2025;33:317–335. doi: 10.1016/j.ymthe.2024.11.028. [DOI] [PubMed] [Google Scholar]
- 3.Friedman K.M., Garrett T.E., Evans J.W., Horton H.M., Latimer H.J., Seidel S.L., Horvath C.J., Morgan R.A. Effective Targeting of Multiple B-Cell Maturation Antigen-Expressing Hematological Malignances by Anti-B-Cell Maturation Antigen Chimeric Antigen Receptor T Cells. Hum. Gene Ther. 2018;29:585–601. doi: 10.1089/hum.2018.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Qin H., Ramakrishna S., Nguyen S., Fountaine T.J., Ponduri A., Stetler-Stevenson M., Yuan C.M., Haso W., Shern J.F., Shah N.N., Fry T.J. Preclinical Development of Bivalent Chimeric Antigen Receptors Targeting Both CD19 and CD22. Mol. Ther. Oncolytics. 2018;11:127–137. doi: 10.1016/j.omto.2018.10.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Roddie C., Lekakis L.J., Marzolini M.A.V., Ramakrishnan A., Zhang Y., Hu Y., Peddareddigari V.G.R., Khokhar N., Chen R., Basilico S., et al. Dual targeting of CD19 and CD22 with bicistronic CAR-T cells in patients with relapsed/refractory large B Cell lymphoma. Blood. 2023;141:2470–2482. doi: 10.1182/blood.2022018598. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Cordoba S., Onuoha S., Thomas S., Pignataro D.S., Hough R., Ghorashian S., Vora A., Bonney D., Veys P., Rao K., et al. CAR T cells with dual targeting of CD19 and CD22 in pediatric and young adult patients with relapsed or refractory B cell acute lymphoblastic leukemia: a phase 1 trial. Nat. Med. 2021;27:1797–1805. doi: 10.1038/s41591-021-01497-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Shalabi H., Qin H., Su A., Yates B., Wolters P.L., Steinberg S.M., Ligon J.A., Silbert S., DéDé K., Benzaoui M., et al. CD19/22 CAR T cells in children and young adults with B-ALL: phase 1 results and development of a novel bicistronic CAR. Blood. 2022;140:451–463. doi: 10.1182/blood.2022015795. [DOI] [PMC free article] [PubMed] [Google Scholar]