The Manufacturing Gap Defining the Next Decade of Cell Therapy

Cell therapy is entering a transition the industry has been quietly bracing for. The first generation of approved products, autologous CAR-T therapies and ex vivo gene-modified cell products, proved the modality works. The next generation is going to look very different, and the manufacturing infrastructure built to support the first wave is not going to carry the second one without significant change.

The shift driving that change is the move toward in vivo cell therapy: engineering immune cells inside the patient's body rather than extracting, modifying, and reinfusing them. Combined with the rise of synthetic biology, therapies built with programmable, multi-gene logic, this represents a qualitatively different class of medicine. And it demands a qualitatively different kind of manufacturing.

Where the First Generation Plateaued

The clinical success of early cell therapies came with a set of well-documented limitations. Autologous CAR-T therapies are remarkably effective in certain hematologic cancers, but they are expensive to manufacture, slow to deliver, and constrained by the need to extract cells from patients who are often already sick. Each dose is essentially a custom product, with a vein-to-vein timeline measured in weeks. The model works clinically, but it does not scale to the patient populations that could benefit.

These limitations are not failures of the modality. They are the natural ceiling of what an autologous, ex vivo manufacturing model can do. To reach broader patient populations, especially in solid tumors, autoimmune disease, and conditions outside major academic centers. The products themselves have to be re-architected.

That re-architecture is happening on two fronts at once. First, allogeneic and off-the-shelf cell therapies are maturing, removing the patient-specific manufacturing constraint. Second, and more transformationally, in vivo cell therapy is emerging: it delivers engineered payloads directly into patients, enabling modifications to occur within the body. Programs from companies pursuing lipid nanoparticle-delivered CAR programming, in vivo gene editing of immune cells, and engineered viral vectors that target specific cell populations are pushing this approach from concept toward the clinic.

Layered on top of both is synthetic biology. Multi-gene constructs that behave less like a static payload and more like a small program. Logic gates that activate a therapy only when specific cellular markers are present. Tunable expression that adjusts dose in response to feedback. Safety switches that allow the therapy to be shut down or eliminated. Researchers and clinicians have described these as "gene circuits," and the analogy that has stuck is software: each functional module behaves like an app, and the therapy as a whole is the operating system.

The Bottleneck Nobody Wants to Own

Here is where the field runs into a structural problem. The contract manufacturing model that supports most of biopharma assumes a clean separation between development and production: a sponsor designs a product, locks the construct, and hands it to a CDMO to manufacture under GMP. That model works reasonably well for monoclonal antibodies and even for first-generation autologous cell therapies, where the construct is relatively simple and the process is well-characterized.

It does not work well for in vivo cell therapy or for synthetic biology more broadly.

When a therapy involves three or four genes packed into a single delivery vehicle, with regulatory elements interacting in ways that affect targeting, expression, stability, and potency, the process and the product are tightly coupled. Optimizing the manufacturing process changes the product. Changing the product forces process redesign. The traditional handoff is to design here, manufacture there, which introduces delays and failures that compound across the development timeline.

In vivo cell therapy intensifies this problem. The delivery vehicle has to find the right cells inside a complex body, deposit a multi-gene payload, and trigger the right behavior; all without the controlled environment of an ex vivo manufacturing suite to verify each step. Characterization, potency assays, and release testing all become harder. The construct and the delivery system have to be co-developed, not designed in sequence.

Sponsors developing these products consistently report the same experience: enormous amounts of time spent on in-house design, followed by stalled tech transfers to CDMOs that lack the platform expertise to engage with the construct itself. The industry has begun to acknowledge, often informally, that "design assistance" has become the scarcest and most valuable service in cell therapy outsourcing, and the one most CDMOs are not structured to provide.

This is the gap that emerging biofabs are trying to fill. Whether it's a former biotech facility spun out into an independent CDMO or a new build from a contract specialist, the common thread is integration: design and GMP manufacturing under one roof, with staff who can engage with the biology, not just the protocol.

What the Next Two Years Look Like

Several shifts are likely to define the field through 2026 and 2027.

In vivo programs will move from preclinical to clinical in meaningful numbers. The first in vivo CAR programs are already in early trials, and the pipeline behind them is substantial. Over the next two years, expect a steady accumulation of clinical data that will start to answer the central questions: Can you reliably engineer the right cells inside the body? Is the safety profile manageable? Does the durability hold up? Each readout will shift industry conviction one way or the other.

Pipeline complexity will keep rising. Programs entering the clinic in the next two years are meaningfully more sophisticated than what came before. Allogeneic platforms, NK cell therapies engineered with multiple functional genes, in vivo gene editing of immune cells, and engineered tissues are all moving past proof-of-concept. Each adds manufacturing demands that don't map cleanly onto existing capacity.

Capacity will continue to be misaligned with demand. The CDMO build-out of the past several years was sized largely for autologous cell therapy work and conventional biologics. Some of that capacity will be repurposed; some will sit underutilized while sponsors with complex programs struggle to find homes for them. Expect more carve-outs, specialty CDMOs, and biotech-to-contract transitions as the market reallocates.

Design-integrated manufacturing will become a recognized category. Today, "design assistance" is a feature some CDMOs claim and few deliver substantively. Over the next two years, expect it to harden into a distinct service tier with its own pricing, contract structures, and quality expectations. Sponsors will increasingly evaluate providers not just on capacity and compliance, but on platform fluency, particularly for synbio and in vivo programs.

Regulatory frameworks will be tested. Multi-gene, conditional therapies and in vivo approaches that engineer cells outside the controlled environment of a manufacturing suite — raise questions about comparability, potency assays, biodistribution, and characterization that current guidance was not written to answer. FDA, EMA, and other regulators are engaged on these questions, but the answers will be shaped in part by the early programs going through review.

The economics of personalization will get pressure tested. Autologous CAR-T has proven clinically valuable but commercially constrained by cost and logistics. The next wave of allogeneic, off-the-shelf, and in vivo approaches informed by synthetic biology platforms, promises more standardized, scalable products. Whether the manufacturing economics actually deliver on that promise will be a defining variable in which programs reach broad patient access.

The Underlying Bet

The investments going into biofabs, design-integrated CDMOs, and synthetic biology platforms reflect a shared bet about where the field is heading; that the future of cell therapy is not just more products like the ones we have, but a qualitatively different class of products that demand a different kind of infrastructure to build. In vivo approaches and programmable gene circuits are the leading edge of that shift.

Whether the bet pays out on a two-year timeline or a five-year one is genuinely uncertain. Clinical programs slip. Manufacturing innovations take longer to validate than anyone expects. The first wave of in vivo and programmable cell therapies will face the same kinds of unexpected setbacks every new modality has faced before it.

But the direction is hard to mistake. The clinical demand for more precise, more controllable, more accessible therapies is real. The scientific tools to build them exist and are maturing quickly. And the manufacturing model required to bring them through development is being assembled now, in real time, by an industry that has finally accepted that the old model isn't going to scale into the next generation.

The companies, sponsors and service providers alike that recognize the shift early will shape the next decade. The ones that don't will spend it catching up.

Informed in part by reporting in Outsourced Pharma, "Gene Circuits And Biofabs: Building The Evolution" by Louis Garguilo, April 1, 2024.