There are two fundamentally different approaches to delivering cell and gene therapy: ex vivo, where a patient’s cells are modified outside the body before being returned, and in vivo, where the therapeutic agent is delivered directly into the body to modify cells in place. The distinction isn’t a minor technical detail — it shapes manufacturing strategy, delivery logistics, targeting precision, and the entire risk profile of a given therapy.
What Distinguishes Ex Vivo From In Vivo Approaches?
In ex vivo therapy, a patient’s own cells — or in some allogeneic approaches, donor cells — are removed from the body, genetically modified or engineered in a controlled manufacturing environment, and then reintroduced into the patient. CAR-T cell therapy is the clearest commercial example: a patient’s T cells are collected, engineered outside the body to express a chimeric antigen receptor targeting cancer cells, expanded to a therapeutic dose, and then infused back into the patient.
In vivo therapy skips this extraction step entirely. The therapeutic agent — typically a viral vector carrying genetic material — is administered directly into the patient’s body, where it must find, enter, and modify the target cells in their natural location. Zolgensma, delivered as a single intravenous infusion that targets motor neuron cells throughout the body, is a widely known in vivo example.
What Is the Central Technical Challenge for Ex Vivo Therapy?
The biggest technical challenge in ex vivo therapy is ensuring that modified cells effectively target only the intended cell population once reintroduced — a property known as transductional efficiency, which measures how precisely a viral vector infects target cells specifically, rather than acting indiscriminately. Achieving high transductional efficiency sometimes requires retargeting: genetically engineering the vector’s surface molecules so that it preferentially binds to and enters only the desired cell type, for instance by inserting a specific ligand onto the vector’s surface to promote interaction with receptors found primarily on target cells.
Using a viral vector with broad tropism — meaning it can transduce a wide range of cell types — is not inherently a technical problem in itself for ex vivo applications, since the modification happens in a controlled setting outside the body. The critical requirement is restricting that broad capability to infect only the specific target cell population once the modified cells are reintroduced, since unintended modification of non-target cells after reinfusion can produce toxicity or unwanted immune responses.
What Is the Central Technical Challenge for In Vivo Therapy?
In vivo therapy faces a related but distinct version of the same targeting challenge, with higher stakes attached to getting it right. Because the therapeutic agent is delivered directly into the body rather than into a controlled external environment, targeting precision has to be achieved in real time, inside a complex biological system the manufacturer cannot fully control during administration. A vector with broad tropism in an in vivo context risks modifying cells throughout the body indiscriminately, which raises both efficacy concerns — diluting the therapeutic effect across unintended tissue — and safety concerns, since off-target genetic modification in healthy tissue carries real risk.
This is part of why in vivo gene therapy development places such heavy emphasis on vector selection and engineering specific to the target tissue: an AAV vector serotype effective for liver-targeted delivery is often a poor choice for delivering the same genetic payload to retinal or central nervous system tissue, since different AAV serotypes have evolved or been engineered to preferentially target different tissue types throughout the body.
How Do Manufacturing and Logistics Differ Between the Two Approaches?
Ex vivo manufacturing is inherently personalized when using a patient’s own cells — each batch is, by necessity, a single patient’s therapy, manufactured to that patient’s specific cell material on a defined timeline between collection and reinfusion. This creates the “vein-to-vein” logistics challenge widely discussed in autologous cell therapy: the elapsed time between collecting a patient’s cells and returning the finished therapeutic product, during which the patient’s underlying disease continues to progress.
In vivo therapy sidesteps this specific logistics challenge, since the therapeutic product itself can typically be manufactured at scale in advance and distributed more like a traditional biologic, without requiring patient-specific manufacturing timing. The trade-off is that in vivo approaches generally cannot be easily reversed or adjusted after administration the way an ex vivo approach — where modified cells can in principle be characterized and quality-tested extensively before ever entering the patient — allows.
Which Approach Is Better Suited to a Given Therapeutic Application?
Neither approach is universally superior; each tends to suit different therapeutic contexts. Ex vivo approaches have proven particularly well suited to applications where modifying a specific, collectible cell population — most commonly T cells or hematopoietic stem cells — addresses the underlying disease mechanism directly, as seen in CAR-T oncology applications and gene-edited hematopoietic stem cell therapies for blood disorders. In vivo approaches have proven better suited to diseases where the target tissue isn’t practically collectible and modifiable outside the body, such as liver-targeted, retinal, or neuromuscular gene therapies, where reaching the affected tissue directly in its natural location is the only practical delivery strategy.
Despite the challenges specific to each approach, significant progress has continued across both ex vivo and in vivo gene therapy development in recent years, with targeting precision, vector engineering, and manufacturing efficiency all improving meaningfully as the field accumulates more clinical and commercial experience with each approach.