Since the approvals of Kymriah (Novartis) and Yescarta (Kite Pharma) in 2017, CAR-T cell therapy has fundamentally reshaped the treatment landscape for B-cell malignancies. The traditional ex vivo CAR-T manufacturing model relies on complex processes involving cell isolation, genetic modification, and expansion. This approach is time-consuming, costly, and highly dependent on individualized cell sources, which creates significant limitations in scalability, accessibility, and global deployment.

To break the limitations of ex vivo CAR-T cells, a novel approach has been established to generate CAR-T cells directly in vivo. Unlike conventional ex vivo methods, this approach delivers CAR genes straight into a patient’s T cells inside the body using targeted delivery systems such as engineered viral vectors, lipid nanoparticles, etc. By avoiding complex manufacturing steps, the in vivo approach can shorten treatment timelines, reduce costs, and minimize risks such as T cell exhaustion and variability caused by extensive ex vivo manipulation. Importantly, in vivo CAR-T technology also enables more precise control of CAR-T cell function through targeted gene delivery to T cells.

In Vivo CAR-T VS Ex Vivo CAR-T

Core Challenges of Ex Vivo CAR-T Therapy

Ex vivo CAR-T therapy has become a milestone in cancer immunotherapy due to its unprecedented efficacy in relapsed or refractory hematologic malignancies. Ex vivo CAR T-cell therapy involves isolating a patient’s or healthy donor’s T-cells, genetically modifying them with a CAR construct designed for specific cancer antigens, then culturing and expanding ex vivo, and finally reintroducing into the patient via infusion to achieve tumor-specific killing.

Figure 1. Process of ex vivo CAR T therapies. Source: reference [1]

Despite its clinical success, this highly individualized treatment model faces major limitations that have hindered its transition from a niche therapy to broad clinical use.

Complex manufacturing and limited access: The ex vivo process includes multiple steps—cell collection, genetic modification, expansion, and reinfusion—and typically takes several weeks. For patients with rapidly progressing disease, these delays can be life-threatening. At the same time, the labor-intensive and customized manufacturing process results in extremely high costs, making access difficult for most patients globally.

Inconsistent efficacy and safety concerns: Variability in donor-derived starting material can significantly influence the final product's quality and characteristics, complicating standardization and quality control. Extended ex vivo culture can also cause T cell exhaustion, reducing their ability to expand and persist in the body. In addition, lymphodepleting chemotherapy is usually required to improve treatment response, increasing toxicity and limiting use in elderly or fragile patients.

Limited expansion beyond blood cancers: While ex vivo CAR-T therapies perform well in hematologic malignancies, their effectiveness in solid tumors remains limited. Major barriers include the suppressive tumor microenvironment, poor T cell infiltration, and tumor antigen heterogeneity.

Key Advantages of In Vivo CAR-T Strategies

In vivo CAR-T cell engineering has emerged to address the fundamental limitations of ex vivo approaches. This strategy aims to engineer a patient’s immune cells directly within their body. In vivo methods use targeted delivery systems such as lipid nanoparticles or viral vectors to generate CAR T cells in situ. This emerging field holds promise to overcome current logistical and cost barriers, enabling faster, scalable, and potentially safer immunotherapies across oncology and autoimmune disease indications.

Simpler treatment and off-the-shelf potential: In vivo CAR-T therapies function as injectable gene medicines that can be produced and stored at scale. Treatment can shift from weeks of hospitalization to a single outpatient injection, enabling faster treatment initiation and significantly improving accessibility.

Lower costs and wider access: By removing expensive steps such as cell isolation, ex vivo expansion, and cold-chain transport, in vivo CAR-T has the potential to reduce overall treatment costs dramatically, making CAR-T therapy more accessible to a broader patient population.

Improved efficacy and safety profile: Engineering T cells within their natural physiological environment helps preserve their native function and proliferative capacity, reducing the risk of exhaustion. Some in vivo platforms may also reduce or eliminate the need for lymphodepleting chemotherapy, lowering toxicity and expanding eligibility. In addition, in vivo approaches can modify multiple immune cell populations, supporting more durable and coordinated antitumor immune responses.

Controlled expression and dosing flexibility: Modern delivery platforms allow precise control over CAR expression levels and duration. For example, lipid nanoparticle–based mRNA delivery enables transient expression for short-term treatment needs, while integrating viral or gene-editing approaches can support long-lasting expression, tailored to the safety and efficacy requirements of different indications.

Vector Platforms for In Vivo CAR Delivery

Currently, two major types of delivery vectors have been engineered to enable in vivo CAR-T cell production: viral vectors ( lentiviral and adeno-associated viral (AAV) systems ) and non-viral vectors (lipid nanoparticle vectors, polymer nanoparticle vectors, and exosomes).

Figure 2. Vector platforms for in vivo CAR delivery. Source: reference [3]

Viral Vectors

Viral vectors are widely used to deliver genes into mammalian cells and form the foundation of in vivo CAR-T cell engineering. The two main viral platforms used for in vivo T cell modification are lentiviral vectors (LVs) and adeno-associated viruses (AAVs). Viral particles are typically produced in HEK-293T cells or dedicated packaging cell lines using transient transfection.

Lentiviral vectors are HIV-1–based systems that integrate transgenes into the host genome, enabling long-term and stable CAR expression. However, native lentiviruses lack T cell specificity in vivo. To achieve targeted delivery, lentiviral envelopes are engineered with T cell–binding ligands—such as CD3-, CD4-, or CD8-specific scFvs—while eliminating native receptor interactions to reduce off-target transduction. These receptor-targeted lentiviruses can directly engage resting T cells without prior activation, achieving efficient and selective gene transfer in vivo. This approach enables direct in vivo generation of functional CAR-T cells, improves transduction efficiency, minimizes off-target effects, and reduces overall treatment cost.

AAV vectors offer an alternative viral platform for in vivo CAR-T generation. AAVs are non-enveloped, single-stranded DNA viruses with a 4.7 kb genome that typically remains episomal in target cells, supporting stable transgene expression without genomic integration. They exhibit low immunogenicity, minimal toxicity, and a strong clinical safety record. T cell targeting can be enhanced through capsid engineering, such as chimeric AAV variants created by combining features from multiple serotypes. These optimized AAVs have demonstrated efficient in vivo T cell transduction and robust antitumor activity in preclinical models, supporting AAVs as a promising and safe option for in vivo CAR-T cell therapy.

Non-Viral Vectors

Non-viral vectors, particularly lipid nanoparticles (LNPs), have become a major focus in in vivo CAR-T development due to their design flexibility, lower risk of insertional mutagenesis, and potential cost advantages. Their success in COVID-19 mRNA vaccines has further validated their clinical potential.

LNPs are nanoscale particles formed by self-assembly of ionizable lipids, helper lipids, cholesterol, and PEGylated lipids. They effectively encapsulate and protect nucleic acids such as mRNA or circular RNA (circRNA) from degradation in the body. For in vivo CAR-T applications, LNPs are surface-modified with targeting ligands—such as antibodies against T cell markers (e.g., CD8)—to enable selective delivery. Once inside T cells, the released nucleic acids are translated into CAR proteins using the cell’s own protein synthesis machinery.

Compared with lentiviral vectors, LNPs offer a stronger safety profile. CAR expression is transient and does not integrate into the genome. After the mRNA degrades, modified T cells can return to their normal state. In addition, LNP-based systems allow dosing flexibility and repeat administration to optimize efficacy while reducing long-term safety risks.

In Vivo CAR-T Research Pipelines

With the ability to bypass ex vivo cell processing and significantly shorten treatment timelines, in vivo CAR-T has become a major focus in cancer immunotherapy research. Most leading programs worldwide are currently in Phase I clinical trials. Lentiviral-based approaches are primarily advancing in hematologic malignancies, while LNP-based platforms are expanding into autoimmune diseases and solid tumors.

Lentiviral Platforms: Clinical Progress in Hematologic Malignancies

EsoBiotec

Using its third-generation lentiviral ENaBL platform, EsoBiotec developed ESO-T01, the first BCMA-targeted in vivo CAR-T therapy to enter clinical testing for relapsed or refractory multiple myeloma (MM). In July 2025, a Phase I investigator-initiated trial conducted with Union Hospital, Tongji Medical College, was published in The Lancet. All four patients treated at the lowest dose level—without lymphodepleting chemotherapy—achieved clinical responses, including two complete response and two partial responses, demonstrating the feasibility of a lymphodepletion-free approach. Safety findings included chills and fever in all patients, with three cases of hypotension requiring medical intervention; all adverse events were manageable. In March 2025, AstraZeneca acquired EsoBiotec for USD 1 billion (USD 425 million upfront plus USD 575 million in milestones), strengthening its presence in hematologic oncology.

Umoja Biopharma

Umoja’s VivoVec platform integrates CD3-targeting antibodies with the CD58 and CD80 costimulatory domains on the lentiviral surface, while also incorporating a Cocal fusion glycoprotein to target the LDL receptor on T cells. This design significantly improves transduction efficiency and T cell activation. The CD19-targeted candidate UB-VV111 received FDA IND approval in August 2024 and has entered a Phase I dose-escalation trial in hematologic malignancies. Umoja is accelerating development through strategic collaborations, including partnerships in 2024 with IASO Biotherapeutics, AbbVie, and Nona Biosciences to integrate CAR constructs, antibody technologies, and indication expansion.

Interius BioTherapeutics

Interius is developing INT2104, a non-replicating, self-activating lentiviral in vivo CAR-T therapy. By incorporating a CD7 antibody fragment into the viral vector, INT2104 enables dual targeting of T cells and NK cells to generate anti-CD20 CAR-T/NK cells for B-cell malignancies. In July 2024, the program received approval from Australia’s TGA to enter Phase I clinical trials, exploring the therapeutic potential of dual immune cell targeting.

LNP Platforms: Innovation in Autoimmune Diseases and Solid Tumors

Capstan Therapeutics

Capstan Therapeutics is a leader in the LNP-based in vivo CAR-T space. The company was co-founded by CAR-T pioneer Carl June and mRNA pioneer Drew Weissman. Its mRNA-tLNP platform uses optimized lipid components—such as the novel ionizable lipid L828—combined with targeting ligands to reduce liver accumulation and enhance immune cell specificity.

Capstan’s lead candidate, CPTX2309, is an anti-CD19 in vivo CAR-T therapy for B cell–mediated autoimmune diseases and entered Phase I clinical trials in June 2025. A study published in Science the same month showed that CD8 antibody–conjugated L829-tLNPs selectively targeted human CD8⁺ T cells. Delivery of anti-CD19 CAR mRNA rapidly generated functional CAR-T cells in humanized mouse models, demonstrating antigen-specific cytotoxicity, proliferation, and cytokine secretion, while potentially reducing the risk of cytokine release syndrome (CRS).

In June 2025, AbbVie acquired Capstan for USD 2.1 billion, gaining access to its tLNP platform and CPTX2309, and further expanding its footprint in autoimmune and oncology research. Meanwhile, companies such as Myeloid are also exploring LNP-based in vivo CAR-T approaches for solid tumors, aiming to overcome the indication limits of traditional CAR-T therapies.

Summary

Although in vivo CAR-T therapies still face challenges—such as delivery efficiency, long-term safety, and effective infiltration of solid tumors—early clinical data have confirmed their feasibility. At the same time, increased investment and acquisitions by major pharmaceutical companies are accelerating clinical translation. Looking ahead, further optimization of delivery platforms, advances in CAR design, and the development of combination strategies may enable in vivo CAR-T therapies to move beyond blood cancers and address a broader range of diseases, establishing a new treatment paradigm for both cancer and autoimmune disorders.

Huateng Pharma provides a comprehensive range of high-quality PEG derivatives for LNP formulations, supporting in vivo CAR-T, mRNA, and gene delivery applications. Our PEG solutions are designed to enhance LNP stability, circulation performance, and formulation consistency, with reliable supply from early research through clinical development.

References:
[1] Advancements and challenges in developing in vivo CAR T cell therapies for cancer treatment

Bui, Thuy Anh et al. eBioMedicine, Volume 106, 105266
[2] Huang, Y., Cao, R., Wang, S., Chen, X., Ping, Y., & Zhang, Y. (2025). In vivo CAR-T cell therapy: New breakthroughs for cell-based tumor immunotherapy. Human Vaccines & Immunotherapeutics, 21(1), 2558403. https://doi.org/10.1080/21645515.2025.2558403
[3] Huang, Y., Cao, R., Wang, S., Chen, X., Ping, Y., & Zhang, Y. (2025). In vivo CAR-T cell therapy: New breakthroughs for cell-based tumor immunotherapy. Human Vaccines & Immunotherapeutics, 21(1), 2558403. https://doi.org/10.1080/21645515.2025.2558403