How In Vivo Car T Therapy Is Engineered?

In vivo CAR T therapy is engineered by delivering genetic instructions directly into a patient’s body so their own T cells — a type of immune cell — are reprogrammed to recognize and attack cancer cells. Instead of removing T cells, modifying them in a lab, and infusing them back, the modification happens inside the body. The delivery tools are typically engineered viruses or lipid nanoparticles that carry genetic payloads into T cells, where the cells then produce a receptor called a CAR, or chimeric antigen receptor, on their surface.

What Makes In Vivo CAR T Therapy Different From Conventional CAR T?

Conventional CAR T therapy is an ex vivo process. That means “outside the body.” It starts with leukapheresis, a procedure where blood is drawn and T cells are separated out. Those cells are then genetically modified in a specialized lab, expanded to large numbers over one to two weeks, and infused back into the patient after lymphodepleting chemotherapy.

In vivo CAR T skips nearly all of that. The genetic payload goes straight into the patient. No cell collection. No lab expansion. No weeks of waiting.

That distinction matters for several reasons. Ex vivo manufacturing is complex, expensive, and not available at every hospital. Some patients deteriorate before their modified cells are ready. In vivo approaches could potentially reduce cost and wait time, though those advantages have not yet been confirmed in large human trials.

There is one clarification worth making. “In vivo CAR T” is not a single technique. It is a category that includes several different delivery strategies, each with its own mechanism and its own stage of development. Some use viral vectors. Some use lipid nanoparticles. Some target T cells directly. Others use a two-step approach. Lumping them together as one method obscures real differences in how they work and how far along they are.

How Is the Genetic Payload Delivered Into T Cells Inside the Body?

Delivery is the central engineering problem. A CAR is a engineered receptor, and the instructions for building it must reach T cells specifically — not muscle, not liver, not every cell in the body. Two delivery platforms dominate current research.

Viral Vectors

Lentiviruses and adeno-associated viruses are the most studied delivery vehicles. These are modified viruses with their disease-causing genes removed and replaced with the CAR gene. The virus naturally enters cells and inserts its genetic cargo into the cell’s DNA or delivers it as a stable episome.

The challenge is targeting. A lentivirus injected into the bloodstream will infect many cell types. Researchers address this by engineering the viral surface proteins — the envelope glycoproteins — so they bind specifically to markers found on T cells, such as CD3, CD4, or CD8. This is called pseudotyping. The goal is a vector that infects T cells efficiently while ignoring other tissues.

Some research groups have reported success with this approach in animal models. Whether the same targeting precision holds in humans remains an open question.

Lipid Nanoparticles

Lipid nanoparticles, or LNPs, are the delivery technology behind several mRNA vaccines. They are tiny fat-based spheres that can carry mRNA or other nucleic acids. When injected, they fuse with cell membranes and release their cargo into the cytoplasm.

For in vivo CAR T, the LNP would carry mRNA encoding the CAR. The T cell’s own machinery would read that mRNA and produce the receptor temporarily. The key advantage is that mRNA does not integrate into DNA, so the modification is transient. The key disadvantage is the same — the effect fades as the mRNA degrades, typically within days.

Targeting LNPs to T cells requires attaching molecules to the nanoparticle surface that bind T cell markers. This is an active area of research. Early animal studies have shown that targeted LNPs can deliver mRNA to T cells and produce functional CAR expression, but the efficiency and durability in humans are not established.

What Happens Inside the T Cell After the Payload Arrives?

Once the genetic instructions enter a T cell, the cell’s own protein-making machinery takes over. If the payload is DNA delivered by a viral vector, the DNA must first enter the nucleus. There, it is transcribed into mRNA. If the payload is already mRNA — as with lipid nanoparticles — it can be translated directly in the cytoplasm without entering the nucleus.

Either way, the end result is the same: the T cell reads the instructions and assembles CAR proteins. Those proteins travel to the cell surface and anchor there.

The CAR itself has a defined structure. The outside portion — the ectodomain — is typically a single-chain variable fragment, or scFv, derived from an antibody. This is the part that recognizes a specific target on cancer cells. A hinge region connects it to the cell membrane. Inside the cell, signaling domains — most commonly CD3-zeta plus one or more co-stimulatory domains like CD28 or 4-1BB — transmit an activation signal when the scFv binds its target.

When the CAR binds its target on a cancer cell, the signaling domains activate the T cell. The T cell then releases cytotoxic proteins that kill the target cell. This is the same mechanism that conventional CAR T cells use. The difference is only in how the CAR got there.

One important distinction between DNA and mRNA delivery is durability. DNA-based CAR expression can persist for the life of the T cell. mRNA-based expression is temporary — days, not months. That difference shapes what each approach might be suited for. A transient CAR might be useful for conditions where prolonged immune activation is not desirable. A persistent CAR might be necessary for long-term cancer surveillance. Neither approach has been proven superior in clinical trials.

What Are the Main Engineering Challenges?

Targeting specificity is the first hurdle. The delivery vehicle must reach T cells and avoid other cells. If a viral vector infects liver cells or other tissues, it could cause unintended genetic modification. If a lipid nanoparticle delivers mRNA to the wrong cell type, that cell might produce a CAR it should not have.

Delivery efficiency is the second. Even if the vehicle reaches T cells, not every T cell will take up the payload. The percentage of T cells that actually get modified — the transduction efficiency — determines how strong the response will be. Low efficiency means fewer functional CAR T cells and a weaker anti-tumor effect.

Control over CAR expression is the third. With ex vivo manufacturing, clinicians know exactly how many CAR T cells they are infusing and how much CAR each cell expresses. With in vivo delivery, that control is reduced. The dose of vector or nanoparticle is known, but the resulting number of modified T cells is not precisely predictable. This makes dosing and safety monitoring more complex.

Immune responses to the delivery vehicle are the fourth. Viral vectors can trigger immune reactions. Lipid nanoparticles can also provoke immune responses, especially with repeated dosing. If the body neutralizes the delivery vehicle before it reaches T cells, the therapy fails.

Finally, there is the question of durability. Some in vivo approaches may produce CAR T cells that persist for months or years. Others may produce only a transient effect. For cancer treatment, persistence is often considered important for preventing relapse. But persistence also carries risks, including prolonged immune activation and potential long-term side effects. The optimal duration of CAR expression for different diseases is not known.

Where Does In Vivo CAR T Research Stand Now?

Most published work on in vivo CAR T is preclinical — meaning it has been done in animals, not humans. Several research groups have demonstrated that viral vectors and lipid nanoparticles can deliver CAR genes to T cells in mice, and that those T cells can kill target cancer cells. Some studies have shown that a single injection can produce enough CAR T cells to shrink tumors in animal models.

Human trials are in early stages. A small number of phase 1 trials have begun or are planned. Phase 1 trials primarily test safety and dosing, not effectiveness. They involve small numbers of participants and are designed to identify whether a treatment causes harm, not whether it works. It is too early to say whether in vivo CAR T will be effective in people.

Several companies and academic centers are working on different versions of the technology. Some focus on viral delivery. Others focus on lipid nanoparticles. Some target CD19, a marker on B cells used in approved CAR T therapies for certain blood cancers. Others target different markers. The field is moving quickly, but no in vivo CAR T therapy has been approved by the FDA.

It is worth noting that conventional CAR T therapy itself is only approved for a limited number of cancers — primarily certain lymphomas, leukemias, and multiple myeloma. Solid tumors have proven much harder to treat with CAR T, in part because the tumor microenvironment suppresses T cell activity and because finding a target that is on cancer cells but not on healthy tissue is difficult. In vivo approaches face the same biological obstacles. Changing the delivery method does not change the biology of the cancer.

What Are the Potential Advantages and Unknowns?

The potential advantages of in vivo CAR T are mostly practical. If it works, it could:

• Eliminate the need for leukapheresis, a procedure that requires specialized equipment and trained staff
• Reduce or remove the weeks-long manufacturing wait
• Lower the cost of treatment, since there is no personalized lab manufacturing step
• Allow treatment to start immediately, which may help patients whose disease is progressing quickly
• Make CAR T therapy available at more hospitals, including those without cell therapy labs

These are plausible advantages, but none have been confirmed in human trials. The actual cost, speed, and accessibility of in vivo CAR T will depend on the delivery platform, the manufacturing of the vector or nanoparticle, and the clinical infrastructure required to administer it.

The unknowns are substantial. It is not known:

• Whether in vivo delivery can achieve the same level of CAR expression as ex vivo manufacturing
• How long CAR T cells produced in vivo will persist in humans
• Whether the immune system will reject the delivery vehicle before it works
• What the side effect profile will look like, including cytokine release syndrome and neurotoxicity — two serious complications seen with conventional CAR T
• Whether repeated dosing is safe or effective
• How in vivo CAR T will perform in solid tumors, where conventional CAR T has had limited success

The honest assessment is that in vivo CAR T is a promising area of research with a sound scientific rationale. But the gap between animal models and human outcomes is wide. Many therapies that work in mice do not work in people. Until controlled human trials report results, the effectiveness and safety of in vivo CAR T remain unproven.