While in vivo CAR-T therapies eliminate complex ex vivo manufacturing, they introduce a new critical variable: the patient's own immune system. The therapy's efficacy relies on modifying T-cells within the body, but each patient's immune status is different, especially after prior treatments. This makes optimizing and standardizing the dose a significant challenge compared to engineered cell therapies.

The success of early CAR-T cell therapies was partly luck. Future therapies face a high bar, as an ideal target must meet three criteria: 1) be abundant on cancer cells, 2) be indispensable for the cancer's survival, and 3) be dispensable for the patient's healthy tissues to avoid lethal toxicity.

While redosing may be an option for RNA-based in vivo CAR-Ts, viral vector-based platforms face a significant challenge. The potential for an immune response against the vector, a well-known issue in AAV gene therapy, could prevent subsequent doses and limit the long-term treatment strategy for these therapies.

An investigational in vivo CAR-T therapy uses viral particles infused directly into the patient to convert their T-cells into CAR-T cells. This approach eliminates the complex steps of apheresis, lymphodepletion, and ex vivo manufacturing, effectively creating an off-the-shelf product that becomes an autologous treatment inside the body.

Developing CAR T-cell therapies for solid tumors is difficult because many tumor-associated antigens are also expressed on normal tissues. This creates a significant risk of "on-target, off-tumor" effects, causing severe toxicity. Mitigating this risk, for instance with engineered "kill switches," is as crucial as preserving the therapy's efficacy.

To increase safety and efficacy, next-generation CAR-T therapies use "logic-gated" designs. These constructs only activate when they detect the co-expression of multiple antigens—a signature unique to tumor cells—thereby avoiding off-target toxicity on healthy tissues that may express only one of the antigens.

Despite excitement for in-vivo CAR-Ts, the high response rates and multi-year survival of current autologous therapies create a significant competitive moat. New modalities must not only match this efficacy but also prove long-term durability, a high bar that insulates incumbents in indications like multiple myeloma for the foreseeable future.

Despite exciting early efficacy data for in vivo CAR-T therapies, the modality's future hinges on the critical unanswered question of durability. How long the therapeutic effects last, for which there is little data, will ultimately determine its clinical viability and applications in cancer versus autoimmune diseases.

Early data from an in vivo CAR-T therapy suggests a paradigm shift is possible. By engineering T-cells directly inside the patient with a simple infusion, this approach could eliminate the need for leukapheresis and external manufacturing, completely disrupting the current cell therapy model.