Introduction
The B-cell maturation antigen (BCMA)-targeted chimeric antigen receptor (CAR)-T cells ciltacabtagene autoleucel (cilta-cel) and idecabtagene vicleucel (ide-cel) have achieved remarkable success in relapsed/refractory multiple myeloma. In terms of antitumor efficacy, cilta-cel has outperformed ide-cel, in real-world comparisons (progression-free survival; p<0.001, HR=0.45, CI 0.36 to 0.63) 1 as well as their respective pivotal trials (overall response rate of 97% for cilta-cel vs 73% for ide-cel). 2 3 Cilta-cel is the only Food and Drug Administration (FDA)-approved CAR-T cell product that uses a nanobody-based binder, which engages two BCMA epitopes; all others use a single-chain variable fragment (scFv)-based binder. Therefore, its superior efficacy has been attributed to the biparatopic binder design, a strategy which has since been pursued for solid tumors. 4
However, cilta-cel and ide-cel also differ in a second, overlooked variable: the gene regulatory elements used to drive CAR transgene expression. The cilta-cel CAR transgene is driven by an elongation factor α (EF1α) promoter stabilized with a woodchuck hepatitis post-transcriptional regulatory element (WPRE). EF1α is a ubiquitously expressed mammalian cellular promoter and a strong constitutive driver of transgene expression that is used in the majority of clinically approved lentiviral CAR-T cell products. 5 – 8 The WPRE is a commonly used element in lentiviral vectors, located at the 3′-untranslated region of the transgene to increase messenger RNA (mRNA) levels by enhancing transcript termination, which ultimately enhances transgene expression. 9 The ide-cel transgene is expressed under the myeloproliferative sarcoma virus enhancer (MND) promoter without WPRE. 10 MND is a gamma retrovirus-derived promoter, thus absent in human cells, and drives high transgene expression with high per copy activity. 5 11 Ide-cel is the only FDA-approved lentiviral CAR-T cell product expressing the CAR under the MND promoter.
The level and kinetics of CAR expression throughout antigen exposure are increasingly recognized as determinants of CAR-T cell function. 12 Gene regulatory elements determine the transcriptional and post-transcriptional control of CAR expression next to posttranslational mechanisms such as internalization, degradation, and recycling of CAR proteins. 13 Therefore, we asked whether the divergent regulatory configurations driving CAR expression in cilta-cel and ide-cel, rather than their different binder platforms to which their efficacy has been attributed, account for their different clinical outcomes. Based on the clinical efficacy data, we hypothesized that the EF1α +WPRE configuration confers superior efficacy over MND -WPRE , for both dual-VHH (dVHH) and scFv binders. To test this, we reconstructed the binder and the regulatory configuration of cilta-cel and ide-cel from their published sequences and expressed each binder under both configurations while holding all other elements identical. We do acknowledge that we are using reconstructed versions of the molecular constructs and are not producing bona fide “ide-cel” or “cilta-cel” products in our experimental approach, but the nomenclature is used for simplicity to describe the transgene designs.
Methods
Cell lines
SUP-T1 (CRL-1942), Jurkat (TIB-152), Hek293T (CRL-3216) and MM.1S (CRL-2974) cell lines were obtained from the American Type Culture Collection, STR-authenticated and Mycoplasma tested. Cells were cultured in RPMI medium (Thermo Fisher Scientific #72400047) supplemented with 10% fetal bovine serum (Life Technologies, #A52567-01) and 1% penicillin/streptomycin (Thermo Fisher Scientific #15140163), referred to as R10, at 37°C, 5% CO₂. The MM.1S cell line was transduced to stably express a click beetle green luciferase-GFP reporter and GFP-sorted for bioluminescence imaging (BLI). A nuclear factor of activated T cells (NFAT)-GFP Jurkat reporter line was generated by lentiviral transduction with the NFAT eGFP Reporter Lentivirus (BPS Bioscience, cat #79922), and selected for transduced cells through puromycin.
Mice
NSG mice (Jackson #JAX:005557) were bred at the MGB Cancer Institute. Experiments followed an Institutional Animal Care and Use Committee (IACUC)-approved protocol #2020N000114. Each in vivo experiment comprised five groups: untransduced T cells (UTD) and each binder (scFv, dVHH) under each configuration (MND -WPRE , EF1α +WPRE ) with n=5 mice per group (25 mice per experiment). Mice at 8–12 weeks were intravenously injected with 1×10 6 MM.1S across all in vivo experiments. Only mice with confirmed tumor engraftment were randomized by BLI tumor burden. CAR-T cells were thawed, rested one hour in R10 with 20 IU/mL interleukin (IL)-2 (PeproTech #200-02) and injected intravenously (curative dose 2×10 6 cells, stress dose 0.5×10 6 cells). Weekly BLI assessments were performed through intraperitoneal injection of D-luciferin by two veterinary technicians blinded to treatment (Thermo Fisher Scientific #88294, 30 mg/mL), imaged with an Ami HT/HTX spectral apparatus and analyzed in Aura V.5.0.0.
Production of lentiviral vectors
Amino acid sequences were codon-optimized for human expression, synthesized (GenScript), and cloned into a third-generation self-inactivating lentiviral vector backbone. Lentiviral particles were produced by transient transfection of HEK293T cells (Lipofectamine 3000; transfer plus pMDLg/pRRE, pRSV-Rev, pVSV-G packaging plasmids; details in online supplemental methods ), concentrated by ultracentrifugation, and stored at -80°C. Vector titers were determined on SUP-T1 cells.
Titration of lentiviral vectors on primary T cells
Leukapheresis products from anonymous healthy human donors (MGH blood bank, IRB-exempt) were used to isolate T cells (Stemcell #15061). T cells were thawed, activated with CD3/CD28 Dynabeads (Life Technologies #40 203D; 3:1 bead-to-T-cell ratio) and plated at 1×10 5 cells/100 µl. On day one, T cells were transduced with the lentivirus across a range of multiplicities of infection (MOI). UTD from matching donors served as controls. Cultures were expanded with fresh R10/IL-2 every two days, with Dynabead removal on day 5, until day 14. mCherry or CAR surface expression was measured by flow cytometry (Miltenyi #130-136-457) on day 14 (gating is described in online supplemental figure S1 ).
Digital droplet PCR of CAR transgene copies
For each construct, three MOIs yielding 30% transduction (±10%) were selected, and genomic DNA was purified from the corresponding bulk (unsorted) populations (5×10 5 cells), including UTDs (DNeasy Blood and Tissue Kit, Qiagen #69506; Qubit High Sensitivity DNA Assay, Thermo Fisher Scientific # Q33230 ). Droplet generation, amplification and reading were performed on Bio-Rad instruments (Automated Droplet Generator, C1000 Touch, QX600). No-template and matched-donor UTD controls were included in every run (see methods in online supplemental appendix ). One MOI per construct was chosen so that bulk vector copy number (VCN) was closely matched across all four constructs.
CAR-T cell production
For functional and in vivo studies, T cells were activated as above (1×10⁶/mL, 24-well) and transduced at the selected MOI titrated to match bulk VCN across constructs. At day 8, CAR+ cells were enriched by BCMA-PE labeling (Miltenyi #130-133-888 #130-048-801) and magnetic column separation (Miltenyi #130-042-401; purity >95% positivity for mCherry and CAR surface expression) and cultured in IL-2 until day 14. CAR-T cells were cryopreserved in FBS+10% DMSO (Thermo Fisher Scientific #85190; see methods in online supplemental appendix ).
NFAT reporter assay
Tonic signaling was assessed in the NFAT-GFP Jurkat reporter cell line, transduced across MOI 0.625–7.5. To compare tonic signaling independently of CAR surface expression level, EF1α +WPRE and MND -WPRE conditions were matched for CAR surface mean fluorescent intensity (BCMA-PE). On day 7 after transduction, in the absence of BCMA + target cells, NFAT-GFP and CD69 expression were quantified by flow cytometry.
Flow cytometry
Cells were stained 30 min at 4°C, resuspended in 4′,6-diamidino-2-phenylindole (DAPI; Miltenyi #130-111-570, 1:100) for live/dead cell discrimination immediately prior to analysis on a NovoCyte Penteon (Agilent). CAR expression was detected using recombinant human BCMA-PE (Miltenyi #130-133-888) or BCMA-APC (Miltenyi #130-136-457). CD69 was detected with an anti-CD69 antibody (clone FN50, BioLegend #310912). Data were analyzed in FlowJo V.10.10.0 (gating is described in online supplemental figure S1 ).
Cytokine analysis
Plasma from tumor-bearing mice (EDTA) was assayed for human cytokines by a bead-based multiplex immunoassay (LEGENDplex Human CD8/NK Panel, BioLegend #741187). Samples were diluted 1:10 in duplicates, acquired on Penteon (Agilent) and concentrations (pg/mL) interpolated from a standard curve using LEGENDplex analysis software.
Results
CAR-T cells driven by MND -WPRE show higher transduction than EF1α +WPRE at matched MOI across both BCMA binders
To dissect the contributions of the gene regulatory configuration (MND -WPRE vs EF1α +WPRE ) and binder architecture (an ide-cel-like scFv vs a cilta-cel-like biparatopic dVHH), we expressed each binder under the two clinically used regulatory configurations. All constructs shared identical transmembrane and intracellular architecture (a CD8 hinge and transmembrane domain, a 4–1BB costimulatory domain, and a CD3ζ intracellular domain) linked to mCherry via a T2A ribosomal skip element. We did not recapitulate their full, proprietary manufacturing and therefore refer to them as “ide-cel-like” and “cilta-cel-like” ( figure 1 A). Following lentiviral vector production in HEK293T cells, vectors were titrated on SUP-T1 cells by transduction rate to define MOIs, and primary human T cells were transduced across an MOI series. Functional titers did not differ between regulatory configurations on SUP-T1 cells ( figure 1 B). On primary T cells, those transduced with the dVHH vector reached lower transduction efficiencies with the EF1α +WPRE cassette than with the MND -WPRE (p=0.0375, figure 1 C), with a similar trend for the scFv constructs. At a matched MOI of 2.5, CAR-T cells expressing either binder under MND -WPRE had higher VCN than those expressing it under EF1α +WPRE (p=0.0010 and p=0.0037 for scFv and dVHH, respectively; figure 1 D). Consistent with this, CAR-T cells driven by MND -WPRE showed higher transduction than those driven by EF1α +WPRE at matched MOI, measured by mCherry ( figure 1 E, online supplemental figure S1 ) and CAR surface expression ( figure 1 F) for both binders.
CAR-T cells driven by MND -WPRE show greater tonic signaling than EF1α +WPRE across both BCMA binders
To assess CAR-T cell state in the absence of target antigen, we expressed the vectors in an NFAT-GFP Jurkat reporter cell line. At matched CAR surface expression ( online supplemental figure S2 ), CAR-T cells driven by MND -WPRE showed higher NFAT-GFP and CD69 expression than those driven by EF1α +WPRE , with overall greater tonic signaling for the dVHH than the scFv constructs ( figure 2 A and B).
CAR-T cells driven by EF1α +WPRE confer superior in vivo tumor control and survival than MND -WPRE across both BCMA binders
To study the impact of the regulatory configuration on CAR-T cell function in primary human T cells without bias from differential integration, we titrated the vectors to matched VCN by ddPCR ( online supplemental figure S3 A). The titration yielded narrow VCN ranges for both binders (0.5–0.9 for scFv and dVHH; online supplemental figure S3 B) with comparable frequencies of mCherry + (30–49% for scFv, 20–47% for dVHH; online supplemental figure S3 C) and CAR + (30–43% for scFv, 23–47% for dVHH; online supplemental figure S3 D) cells. We then treated MM.1S-bearing mice with a curative dose of 2×10⁶ CAR-T cells ( figure 3 A). By BLI, CAR-T cells driven by EF1α +WPRE cleared tumor across both binder platforms, whereas MND -WPRE -driven CAR-T cells did not ( figure 3 B and C). Consistent with this, mice treated with EF1α +WPRE -driven CAR-T cells had higher plasma levels of effector cytokines, such as interferon-γ, perforin, granzyme A and B, at day seven than those treated with MND -WPRE -driven CAR-T cells ( figure 3 D). For the stress model of MM.1S-bearing mice, the CAR-T cell population was further enriched for the CAR+ fraction through column-based sorting with narrow VCN ranges in the CAR + -enriched population (1.3–2.3 for scFv, 1.4–1.9 for dVHH; online supplemental figure S4 A-C). To increase stringency, the dose was lowered to 0.5×10⁶ CAR-T cells ( figure 4 A). Again, CAR-T cells driven by EF1α +WPRE outperformed those driven by MND -WPRE , regardless of the binder ( figure 4 B). This translated into prolonged survival of mice treated with EF1α +WPRE -driven CAR-T cells compared with mice treated with MND -WPRE across both binders ( figure 4 C).
Discussion
In reconstructed CAR vectors modeling the two clinically approved BCMA-targeted products ide-cel (scFv expressed under MND -WPRE ) and cilta-cel (dVHH expressed under EF1α +WPRE ) the regulatory configuration determined antitumor efficacy independently of the binder, and may therefore contribute to the efficacy difference between ide-cel and cilta-cel. We observed that CAR-T cells driven by MND -WPRE showed higher transduction rates and greater tonic signaling than CAR-T cells driven by EF1α +WPRE . Furthermore, MND -WPRE -driven CAR-T cells showed inferior effector function independent of the antigen-binding domain. Thus, the superior efficacy of the cilta-cel-like construct is abrogated when its biparatopic binder is expressed under MND -WPRE , and the ide-cel-like scFv achieved greater antitumor activity under EF1α +WPRE than it did not reach under MND -WPRE . Because functional comparisons in vivo were performed at matched VCN and proportion of CAR positive cells, differences in outcome cannot be explained by differences in vector integration or transduction rate. To our knowledge, this is the first evidence that the regulatory configuration can determine antitumor efficacy independently of the binder design and may therefore contribute to the differences between the two approved BCMA-targeted products, which has so far been attributed to cilta-cel’s biparatopic binder.
Our in vitro findings are consistent with previous reports that MND drives higher titers 14 and transduction rates 14 15 than EF1α in scFv-based CD19 CAR-T cells. These metrics capture how efficiently a vector can be manufactured, but not necessarily how well CAR-T cells perform.
Others have shown that a high density of CAR molecules promotes receptor clustering, tonic signaling and inferior antitumor effects. 16 Accordingly, in BCMA-targeted scFv CAR-T cells, high CAR surface density has been associated with reduced cytotoxicity in vitro and in vivo and with transcriptomic and epigenetic signatures of tonic signaling and exhaustion. 17 Furthermore, transient rest from CAR signaling can prevent or reverse exhaustion and restore the antitumor functionality. 18 These mechanisms cannot explain the greater tonic signaling of the MND -WPRE compared with EF1α +WPRE constructs we observed here, as CAR surface density was matched between regulatory element configurations in the NFAT reporter cell line, and the CAR itself (hinge, transmembrane and costimulatory domains) was identical. CAR surface density is a steady-state measure, set by the relation of synthesis, internalization and degradation, whereas NFAT and CD69 signaling accumulates over time. In addition to steady-state CAR density, as assessed here, the temporal behavior of CAR expression might matter as well. Consistent with a potential value of lower or dynamically regulated CAR expression, Eyquem et al placed the CAR under control of the endogenous T cell receptor constant alpha chain locus (rather than a constitutive promoter such as EF1α), which lowered CAR levels on target binding, reduced exhaustion and improved efficacy compared with EF1α-driven CAR-T cells. 19 Similarly, expression of an scFv-based CD19 4-1BB CAR under the Wiskott-Aldrich syndrome promoter in lentiviral vectors reproduced TCR-like expression kinetics, although tumor control did not differ from EF1α-driven CAR-T cells in a Burkitt lymphoma model. 20 Whether gene regulatory elements permit such modulation depends on its responsiveness to activation-associated transcription factors, and synthetic promoters built from defined transcription factor binding sites can be engineered to couple CAR expression to T-cell state. 21
Our in vivo findings appear to contradict studies in CD19-targeted scFv-based CAR-T cells, where MND-driven CAR-T cells outperformed EF1α at borderline significance in a Raji model 14 and more clearly in a Nalm-6 model. 15 This may reflect a CAR-T cell manufacturing bias, as neither study reports normalization to VCN, so that differences in transduction may have confounded the comparison. Alternatively, the differences could be target antigen related, since CAR-T cell activation has been shown to be antigen density-dependent. 22 23 CD19 is expressed at high, homogeneous density on B-lineage blasts, 24 whereas BCMA is expressed at lower, more heterogeneous density and is subject to shedding. 25 Thus, regulatory configurations may be differentially favorable by antigen level, where MND-driven CAR-T cells might be better suited to higher, and EF1α +WPRE -driven cells to lower target antigen expression levels. Given this discrepancy between MND-driven CARs when targeting CD19 versus BCMA, it remains unclear how these gene regulatory elements affect CAR-T cell responses in solid tumors, where antigen density might be more heterogeneous and CAR-T cells face chronic antigen exposure and a suppressive tumor microenvironment. 4 Furthermore, all constructs used here contain a 4-1BB domain and costimulatory domains differ in their downstream signaling cascades with CD28 driving greater signaling per receptor molecule. Thus the effect for 4-1BB CARs observed here cannot be assumed to hold for CD28-based products. 16
Our study has several limitations: We compared the two clinically deployed configurations as units and did not isolate the contribution of the individual elements; the promoter effect (MND vs EF1α), the presence of WPRE and the difference in vector length. Each of these factors may contribute to the efficacy, such as WPRE increases transgene mRNA and expression, 6 and their roles in the observed differences remain to be defined. As tonic signaling was assessed only in an NFAT-GFP Jurkat reporter line as a surrogate readout but not in primary human CAR T-cells, we cannot establish a causal link between the greater tonic signaling of MND -WPRE and its inferior in vivo efficacy. In addition, xenograft models were performed with one tumor cell line, so it remains unclear whether the superiority of EF1α +WPRE is maintained at different antigen densities. Furthermore, xenograft models in immunodeficient mice do not capture CAR-T cell persistence and memory formation, which are established drivers of durable treatment responses in patients, and the distinct contribution from the biparatopic binding design compared with the scFv remains unclear. Finally, we reconstructed the molecular design of the clinically approved constructs but not their proprietary manufacturing processes.
Our results raise the possibility that the inferior efficacy of ide-cel compared with cilta-cel in pivotal trials and real-world data may be attributable, in part, to the choice of MND -WPRE rather than EF1α +WPRE for CAR transgene expression, although this remains to be confirmed for the clinical products. Regulatory elements are selected based on metrics used to benchmark vectors such as titer, transduction, and VCN and may be chosen deliberately to maximize CAR transgene expression. However, this might favor the configuration that promotes inferior effector function. Although MND -WPRE gave higher titers, transduction rates and VCN, it conferred inferior tumor control in vivo at matched VCN. Therefore, regulatory elements may be better selected by the antitumor efficacy they confer rather than manufacturing parameters alone.
Supplementary material
Acknowledgements
VB is supported by the German Research Foundation (Deutsche Forschungsgemeinschaft, BL 2044/1-1). SP is supported by NIH K99CA303346. MBL is supported by 1K08CA289948-01. MVM is supported by NIH R01CA249062 and NIH R01CA238268 and Parker Institute for Cancer Immunotherapy (PICI).
Provenance and peer review: Not commissioned; externally peer reviewed.
Patient consent for publication: Not applicable.
Ethics approval: Not applicable.
References
1 Hansen DK Peres LC Dima D et al Comparison of Standard-of-Care Idecabtagene Vicleucel and Ciltacabtagene Autoleucel in Relapsed/Refractory Multiple Myeloma J Clin Oncol 2025 43 1597 609 10.1200/JCO-24-01730 PMC12037312 39965175
2 Martin T Usmani SZ Berdeja JG et al Ciltacabtagene Autoleucel, an Anti-B-cell Maturation Antigen Chimeric Antigen Receptor T-Cell Therapy, for Relapsed/Refractory Multiple Myeloma: CARTITUDE-1 2-Year Follow-Up J Clin Oncol 2023 41 1265 74 10.1200/JCO.22.00842 PMC9937098 35658469
3 Munshi NC Anderson LD Jr Shah N et al Idecabtagene Vicleucel in Relapsed and Refractory Multiple Myeloma N Engl J Med 2021 384 705 16 10.1056/NEJMoa2024850 33626253
4 Escobar G Berger TR Maus MV CAR-T cells in solid tumors: Challenges and breakthroughs Cell Rep Med 2025 6 102353 10.1016/j.xcrm.2025.102353 PMC12711667 40945517
5 Koldej RM Carney G Wielgosz MM et al Comparison of insulators and promoters for expression of the Wiskott-Aldrich syndrome protein using lentiviral vectors Hum Gene Ther Clin Dev 2013 24 77 85 10.1089/humc.2012.244 PMC4003467 23786330
6 June Carl H Levine Bruce L Porter David L et al Use Of Chimeric Antigen Receptor-Modified T Cells To Treat Cancer Univ Pennsylvania 2012 WO
7 Tina A Brian C Randolph GJ et al Articles of manufacture and methods for treatment using adoptive cell therapy Juno Therapeutics Inc 2022 US
8 Enrique ZU Bcma-Targeted Car-T Cell Therapy Of Multiple Myeloma Janssen Research & Development Llc, Janssen Biotech Inc 2021 US
9 Schambach A Galla M Maetzig T et al Improving transcriptional termination of self-inactivating gamma-retroviral and lentiviral vectors Mol Ther 2007 15 1167 73 10.1038/sj.mt.6300152 17406345
10 Quigley Travis RR ANTI - bcma car t cell compositions bluebird bio inc, 2seventy bio inc 2020
11 Montini E Naldini L Booth C et al Balancing efficacy and safety in lentiviral vector-mediated hematopoietic stem cell gene therapy Mol Ther 2025 33 6 8 10.1016/j.ymthe.2024.12.028 PMC11764623 39729983
12 Frigault MJ Lee J Basil MC et al Identification of chimeric antigen receptors that mediate constitutive or inducible proliferation of T cells Cancer Immunol Res 2015 3 356 67 10.1158/2326-6066.CIR-14-0186 PMC4390458 25600436
13 Hinckley-Boned A Barbero-Jiménez C Tristán-Manzano M et al Tailoring CAR surface density and dynamics to improve CAR-T cell therapy J Immunother Cancer 2025 13 e010702 10.1136/jitc-2024-010702 PMC12049969 40300856
14 Ho J-Y Wang L Liu Y et al Promoter usage regulating the surface density of CAR molecules may modulate the kinetics of CAR-T cells in vivo Mol Ther Methods Clin Dev 2021 21 237 46 10.1016/j.omtm.2021.03.007 PMC8027690 33869653
15 Zhang X He X Zhang Y et al Targeting CD19 CAR-T with MND promoter enhances tumour killing J Cell Mol Med 2025 29 e70843 10.1111/jcmm.70843 PMC12439678 40956647
16 Gomes-Silva D Mukherjee M Srinivasan M et al Tonic 4-1BB Costimulation in Chimeric Antigen Receptors Impedes T Cell Survival and Is Vector-Dependent Cell Rep 2017 21 17 26 10.1016/j.celrep.2017.09.015 PMC5645034 28978471
17 Rodriguez-Marquez P Calleja-Cervantes ME Serrano G et al CAR density influences antitumoral efficacy of BCMA CAR T cells and correlates with clinical outcome Sci Adv 2022 8 eabo0514 10.1126/sciadv.abo0514 PMC9524842 36179026
18 Weber EW Parker KR Sotillo E et al Transient rest restores functionality in exhausted CAR-T cells through epigenetic remodeling Science 2021 372 eaba1786 10.1126/science.aba1786 PMC8049103 33795428
19 Eyquem J Mansilla-Soto J Giavridis T et al Targeting a CAR to the TRAC locus with CRISPR/Cas9 enhances tumour rejection Nature 2017 543 113 7 10.1038/nature21405 PMC5558614 28225754
20 Tristán-Manzano M Maldonado-Pérez N Justicia-Lirio P et al Physiological lentiviral vectors for the generation of improved CAR-T cells Mol Ther Oncolytics 2022 25 335 49 10.1016/j.omto.2022.05.003 PMC9163403 35694446
21 Zhang Z Qiu X Ning H et al SPECIFIC: A systematic framework for engineering cell state-responsive synthetic promoters reveals key regulators of T cell exhaustion Quant Biol 2025 13 e97 10.1002/qub2.97 PMC12806122 41675505
22 Ramakrishna S Highfill SL Walsh Z et al Modulation of Target Antigen Density Improves CAR T-cell Functionality and Persistence Clin Cancer Res 2019 25 5329 41 10.1158/1078-0432.CCR-18-3784 PMC8290499 31110075
23 Watanabe K Terakura S Martens AC et al Target antigen density governs the efficacy of anti-CD20-CD28-CD3 ζ chimeric antigen receptor-modified effector CD8+ T cells J Immunol 2015 194 911 20 10.4049/jimmunol.1402346 25520398
24 Pillai V Muralidharan K Meng W et al CAR T-cell therapy is effective for CD19-dim B-lymphoblastic leukemia but is impacted by prior blinatumomab therapy Blood Adv 2019 3 3539 49 10.1182/bloodadvances.2019000692 PMC6880911 31738832
25 Knudsen NH Escobar G Korell F et al In vivo CRISPR screens identify modifiers of CAR T cell function in myeloma Nature 2025 646 953 62 10.1038/s41586-025-09489-8 PMC12806180 40993381
Data availability statement
Data are available upon reasonable request.