In-Vivo Induced CAR-T Cell for the Potential Breakthrough to Overcome the Barriers of Current CAR-T Cell Therapy

Purpose: CAR-T cell therapy has shown impressive success in hematological malignancies, but systemic toxicity (CRS, ICANS, on-target/off-tumor effects) and the complex, costly, individualized manufacturing process of autologous CAR-T cells hinder broader application. Universal allogeneic CAR-T cells have encountered safety concerns, with FDA halting some clinical trials. There is an urgent need for new strategies to overcome these barriers. ---

Hypothesis: No formal testable hypothesis is proposed. The central thesis is: if nanocarriers loaded with CAR genes and gene-editing tools are delivered in vivo to program autologous T cells in situ, then CAR-T cell therapy can be achieved with reduced systemic toxicity, simplified manufacturing, and avoidance of allogeneic safety concerns. ---

Aims: - Review CAR structure design and the five generations of CAR evolution. - Summarize barriers to current CAR-T cell therapy (toxicity, immunosuppressive TME, complex manufacturing). - Review gene-editing tools (ZFN, TALEN, CRISPR-Cas9) used in CAR-T cell therapy. - Review gene delivery systems (viral and non-viral) applied in immunotherapy. - Discuss in vivo CAR-T cell induction as a potential breakthrough, including recent preclinical successes with polymer nanoparticles, lentivirus, and AAV. - Provide future perspectives on in vivo induced CAR-T cells as standard immune-cell therapy. ---

Delivery system: CAR structure (modular): - Extracellular antigen-binding domain (scFv), hinge, transmembrane domain, intracellular signaling domain (CD3ζ + co-stimulatory domains such as CD28 or 4-1BB). CAR generations: - 1st: CD3ζ only; 2nd: CD3ζ + one co-stimulatory domain; 3rd: CD3ζ + two co-stimulatory domains; 4th: TRUCKs (with cytokine/regulatory tools); 5th: antigen-dependent activation control receptor. Gene-editing tools: - ZFN, TALEN, CRISPR-Cas9. Gene delivery systems: - Viral vectors: AAV, adenovirus, lentivirus, bacteriophage. - Non-viral — physical: Electroporation, needle injection, laser irradiation, gene gun. - Non-viral — chemical: Liposomes, cationic polymers, gold nanoparticles, silica nanoparticles, quantum dots, carbon nanotubes, exosomes, ferritin, cell membranes. In vivo CAR-T induction platforms: - Polymeric nanoparticles (PBAE): Poly(β-amino ester) core assembled with CAR DNA or mRNA; exterior with polyglutamic acid (PGA) conjugated to anti-CD3 antibody. Delivered second-generation CD19-specific CAR. - Lentiviral vectors: Encapsulated second-generation anti-CD19 CAR gene; bispecific binder-redirected lentivirus for T cell specificity. - AAV vectors: Encoding third-generation CAR gene. Payloads: CAR plasmid DNA, CAR mRNA, gene-editing tools (CRISPR-Cas9 components). Target antigens discussed: CD19, CD20, CD30, CD37, CD79b, CD123, BCMA, Lym-1, and others. ---

Approach: Narrative review of preclinical and clinical literature. No primary experimental groups. Model systems discussed include: - In vivo: Immunodeficient NOD-scid-IL2Rγ^null (NSG) mice; mouse leukemia models; humanized mouse models. - Clinical: FDA-approved CAR-T products (Kymriah, Yescarta, Tecartus, Breyanzi, Abecma); clinical trials registered on ClinicalTrials.gov (CRISPR-Cas9 edited CAR-T targeting >21 antigens). - Disease context: Hematological malignancies (B-ALL, LBCL, MCL, DLBCL, multiple myeloma), solid tumors. - No primary experimental data in this review. ---

Key methods: Techniques and endpoints highlighted across cited studies: - Flow cytometry for CAR expression on T cells, NK cells, NKT cells. - Antitumor efficacy in mouse leukemia models (tumor regression, survival). - Cytokine release assessment (CRS monitoring). - Gene-editing efficiency (CRISPR-Cas9 target antigen editing, PD-1 knockout). - Biodistribution and transfection specificity. - Pie chart analysis of gene delivery systems from PubMed (18,968 papers; keywords: vector name + gene delivery or CAR gene). ---

Key results: - In vivo CAR-T induction with PBAE nanoparticles: Matthias Stephan's team achieved stable and transient expression of CD19-specific CAR in T cells via CAR-DNA and CAR-mRNA nanoparticles, respectively. Antitumor efficacy comparable to conventional lab-manufactured CAR-T cells without systemic toxicity. - Lentiviral in vivo CAR-T: Buchholz and colleagues induced in situ CAR-T cells in NSG mice with antitumor activity, but observed CRS and unexpected CAR-positive NK and NKT cells due to non-specific lentiviral vector tropism. - Bispecific binder-redirected lentivirus: Lai et al. achieved in vivo specific CAR-T engineering with antitumor activity but relatively low numbers of CAR-expressing T cells. - AAV in vivo CAR-T: Wu et al. showed AAV encoding third-generation CAR sufficiently reprogrammed immune effector cells to generate in vivo CAR-T cells; non-specificity of AAV carrying the CAR gene remains a concern. - FDA-approved CAR-T products: Five approved 2017–2021 (Kymriah, Yescarta, Tecartus, Breyanzi, Abecma). Severe/life-threatening CRS occurs in up to 25% of patients. - CRISPR-Cas9 clinical trials: >21 target antigens registered; CD19 and BCMA account for nearly half. - Gene delivery pie chart: 18,968 papers; viral vectors (AAV, adenovirus, lentivirus, bacteriophage) and non-viral physical/chemical systems represented. ---

Interpretation: The authors conclude that in vivo induced CAR-T cells can potentially overcome current barriers: (1) reduce systemic toxicities (CRS, ICANS) through tumor in-situ editing and expansion of T cells; (2) incorporate cytokine genes into CAR structures to flush the immunosuppressive TME; (3) load gene-editing tools to knock out immune checkpoint genes and reverse T cell exhaustion; (4) resolve difficulty of process standardization and scale-up. The final gene-editor nanoparticles can be produced, stored, and delivered as usual medicines. In vivo induced CAR-T cells are expected to replace current CAR-T cell therapy and become the standard immune-cell therapy for cancers. ---

Limitations: - Review, not primary study: No original experimental data; synthesizes published literature. - No reports of CRISPR-based in vivo CAR-T generation: Despite many in vivo CRISPR gene-editing studies, none have yet generated in vivo CAR-T cells. - Viral vector non-specificity: Lentiviral and AAV vectors have non-specific tropism; CAR-positive NK/NKT cells detected unexpectedly. - Random gene insertion risk: Viral vectors carry universal safety concern of random chromosomal insertion. - Low CAR-expressing T cell numbers: Bispecific binder-redirected lentivirus yielded relatively low numbers of CAR-expressing T cells; toxicity was not evaluated. - Clinical monitoring gap: In vivo editing and expansion status of T cells requires more efforts to monitor. - Manufacturing/regulatory: Optimal delivery systems are still unrealized for clinical use; regulatory approval of nanomedicines remains complex. - Allogeneic CAR-T safety concerns: FDA halted universal CAR-T trials from Allogene due to safety concerns; death case in UCARTCS1A trial.