In vivo manufacturing: when the body becomes the bioreactor

When the breakthrough becomes the bottleneck

The traditional manufacturing model for therapeutic monoclonal antibodies (mAbs)—cell culture in massive bioreactors, multistep purification, the attendant process complexity and quality control required—has produced remarkable medicines. But it has also produced a clear view of its own limitations. Globally, access to these medicines is uneven, owing largely to cost and geographic availability—the infrastructure and complex manufacturing processes required to make such therapies just aren’t feasible in many parts of the world (1).

For advanced therapies such as chimeric antigen receptor (CAR) T cell therapies, the reality is even harsher. The cost of CAR T therapy is staggering, and the treatment and manufacturing processes are complex, time-consuming, and resource-intensive (2). Manufacturers are finding ways to bring the costs down, but these challenges have profoundly limited the scalability of cell therapies thus far. In the US, for example, only 20% of lymphoma patients who are eligible for CAR T therapy actually receive it. This is due to a combination of factors, including cost and regional availability (2).

Beyond cost and access is the question of reach. Biologics, cell therapies, and gene therapies have been life-changing, and life-saving, for millions. But they have potential side effects, and they don’t work for everyone. In some cases, the benefits of a therapy may not justify the cost of making and administering it. Many mAb therapeutics require intravenous administration, with infusion times ranging from 30 minutes to several hours depending on the product, dose, and patient tolerance. This administration adds to the total healthcare costs and affordability and also limits the overall patient experience. Other protein formats, such as smaller antibody fragments and intracellular antibodies, have powerful potential in treating difficult drug targets, but are shown to break down too quickly in the body to be effective.

The industry has worked to mitigate these issues through various means, including longer half-life variants and high-concentration formulations for subcutaneous administration, which can reduce dosing frequency and enable at-home or outpatient administration. But barriers to access remain.

What if there were a way to tackle several of these challenges at once? What if it were possible to manufacture some medicines more effectively, improve patient access and the patient’s treatment experience, and use the same mechanism to treat currently untreatable diseases?

Be your own bioreactor: the outside-in model

The limitations of conventional biologics manufacturing—high costs, complex logistics, and uneven global access—are part of what’s driving researchers to lean into a different approach: in vivo therapies that program the body’s own cells to produce therapeutic molecules. Instead of producing therapeutic proteins or cell therapies in external bioreactors, the body itself becomes the bioreactor.

This idea isn’t new. It’s the basis of approved gene therapies and of mRNA vaccines like the ones that helped curb the COVID-19 pandemic. But a convergence of developments in the past several years has made it newly feasible as a way to mitigate some barriers to access while expanding the range of diseases clinicians can treat. As Scott Ripley, vice president of nucleic acid therapeutics and nanomedicine at Cytiva, explains: “The [body-as-bioreactor] theory was always there. It wasn't a could you, it was a why would you? Or why should you? And what we're starting to see now are very credible reasons why it is sensible.”

It’s all in the delivery

A key element in this convergence is the ongoing maturation of viral vectors and of lipid nanoparticle (LNP) delivery systems like the ones used for COVID vaccines. The rapid global development and rollout of these vaccines offered a powerful testament to the viability of this approach, and RNA-LNP technology, in particular, has continued to mature in the years since.

On the payload side, mRNA development timelines are much faster than those for antibody therapeutics, enabling faster progression from concept to clinical material. On the delivery side is the modularity of LNPs. LNP formulations can be engineered to deliver various nucleic acid payloads, with expression duration primarily determined by the payload.

Ripley describes this as “tuning” the LNP, looking forward to a day when scientists can tweak the LNP to answer a range of questions: “Do you want expression in a certain cell type? Do you want long-lived expression? Do you want short-lived expression? Do you want a burst of activity? Do you want an immunogenic response? Do you want a silencing response?”

LNP technology also opens possibilities that are difficult to reproduce through conventional external manufacturing. Commercial mAb manufacturing typically employs a single cell line expressing a defined protein product—though bispecific antibody formats can be produced from engineered cell lines expressing multiple chains simultaneously. In vivo expression could change that design space. Ripley points to oncology efforts in which multiple mRNA strands are introduced at once (via LNP, for example), yielding several proteins in the same patient. “It opens up a new toolbox in terms of different proteins that might have been cleared from the body quickly,” he says. “[You could have] the ability to get in complex protein combinations that perhaps would have been inefficient or uneconomical to make outside the body previously.”

To be clear, the validation and realization of these capabilities is still a work in progress. LNPs tend to accumulate in the liver, and tissue and cell-type targeting are an area of active research. Moreover, the in vivo delivery of multiple mRNA sequences still faces substantial challenges, including control of dosing and expression. But as the potential benefits come increasingly into focus, so is the drive to overcome such challenges.

In vivo CAR T: the clearest case for why

The appeal of in vivo therapies is starkly illustrated in CAR T cell therapy. CAR T cell therapy has proven extremely effective in treating some B cell malignancies, but the complexities of manufacturing it ex vivo are formidable. Cells must be collected from each patient, engineered outside the body, expanded in a bioreactor, tested, and reinfused. Patients, in the meantime, must undergo chemotherapy to prepare their systems for the treatment and then have to remain near the hospital for monitoring for several weeks after infusion. These logistical complexities have severely restricted the treatment’s scalability such that it’s available mainly to patients in developed countries who can afford it and live near a treatment center (2).

An in vivo approach has the potential to remove some of these barriers. Instead of removing and engineering a patient’s cells, in vivo therapy would use a viral vector or LNP to deliver the “ingredients” to cells to develop cancer-fighting properties where they are. This could bypass the complexities associated with ex vivo cell therapy, including removal and processing as well as lymphodepletion.

Indeed, in vivo CAR T therapy is already on its way to becoming a clinical reality. In early 2026, Kelonia reported FDA clearance of its investigational new drug application for KLN-1010, a lentiviral-based in vivo CAR T cell therapy for multiple myeloma (3). Several in vivo CAR therapies using viral vectors or LNPs are now in Phase 1 trials for cancers as well as for autoimmune disease (4).

Reality check: key challenges ahead

The potential for in vivo therapies is clear, and the work scientists have done over the past decade has made them more feasible than ever. But there are still plenty of issues to resolve. These include:

• Delivery precision: Can these tools reliably target specific cell types without off-target expression?
• Expression control: How can clinicians ensure expression levels fall within the therapeutic window—not too low to be ineffective, not too high to cause toxicity—when we can't directly measure or adjust production post-administration?
• Regulatory adaptation: Current frameworks assume you can characterize and test the product before it enters the patient. In vivo manufacturing inverts this. How can regulators establish acceptable surrogate markers and post-administration monitoring protocols?
• Cost and access: Will the streamlined manufacturing potential translate to real savings for patients and healthcare providers? What do regulators, governments, industry, and researchers need to do to democratize access?

These challenges are substantial, but they aren’t insurmountable. For his part, Ripley is optimistic. “The way that I look at the creativity that's coming through in the nucleic acid space, is it likely that every single one of these mechanisms will make it to market? Certainly not,” he says. “But what we start to see is this toolbox, where it can be used…as a recombinant vaccine in the example of personalized cancer vaccines, it can be used to replace natural proteins, it can be used to introduce recombinant proteins.”

Democratizing development

Just as mAbs aren’t a replacement for small-molecule drugs, the body-as-bioreactor approach won’t replace traditional bioprocessing or cell therapy—instead, it is poised to become a powerful supplement to it. And science has been here before: 30 years ago, a convergence of platform maturity, clinical validation, and demand moved mAb therapeutics firmly into the mainstream. A similar energy is now building behind in vivo approaches. But where the complexity of producing mAbs and cell therapies has limited the reach of these products, Ripley envisions the next revolution as having the opposite effect. A future where cutting-edge technology makes the products more accessible, easier to manufacture, and easier for scientists to develop. “If you get to the universal ability to produce an mRNA sequence and introduce that into an LNP, if you democratize that capability, anybody can sequence anything into that mRNA and it starts to become a little bit simpler to introduce these.” Are we there yet? Not quite. But the journey is under way, and the science behind it is accelerating all the time.