A manufacturing perspective from Stephen Madaras, Business Development Manager, Miltenyi Bioindustry , USA
In vivo gene therapy is moving quickly from scientific promise to practical application. From our vantage point as a manufacturing partner working closely with developers of in vivo gene therapies - including emerging in vivo CAR approaches - the conversation has shifted noticeably over the past few years. Early questions about feasibility and biological rationale are giving way to more pragmatic ones: how to manufacture these vectors reproducibly, define and control potency, and maintain development momentum as programs scale.
In vivo approaches offer clear conceptual advantages over ex vivo therapies. Eliminating leukapheresis, centralized cell processing, and complex vein-to-vein logistics could shorten timelines and simplify supply chains. In practice, however, removing ex vivo control steps does not remove complexity. It relocates it - squarely into vector performance, manufacturing consistency, and chemistry, manufacturing, and controls (CMC) strategy, often much earlier than programs initially expect.
Across in vivo gene therapy programs, a common inflection point occurs as projects move beyond early proof-of-concept. At this stage, the demands placed on the gene transfer vector increase significantly compared with ex vivo applications. Transduction must occur in heterogeneous, often non-activated target cell populations, under variable immune pressure, and without the ability to control exposure, remove excess vector, or select modified cells.
As a result, relatively small differences in vector quality can have outsized consequences in vivo. We routinely see vector specificity, functional potency, and lot-to-lot consistency emerge as central determinants of clinical feasibility. Attributes that might be tolerated – or actively managed – during ex vivo manufacturing can become limiting factors once the vector is administered systemically or locally to patients.
Immunogenicity is another area where theoretical considerations quickly become operational realities. Innate immune activation, adaptive neutralization, and constraints on re-dosing are not abstract risks; they directly influence the effective dose, the durability of response, and the clinical design. Importantly, these biological effects are tightly linked to manufacturing decisions. Particle composition, impurity burden, formulation, and dose definition all shape a product's in vivo behavior.
One recurring pattern we observe is the temptation to defer detailed CMC strategy until after initial clinical signals are established. For in vivo gene therapies, this approach rarely holds. Defining potency without traditional ex vivo functional readouts, characterizing heterogeneous particle populations, and justifying dose based on clinically relevant activity are challenges that surface early – and tend to intensify rather than resolve over time.
From our regulatory interactions, expectations are clear: in vivo gene therapy products must demonstrate robust control of identity, purity, and strength from the early stages of development onward. Programs that invest early in analytical maturity and manufacturing control are better positioned to maintain pace as clinical studies expand and scale-up becomes unavoidable.
These experiences have shaped how we approach in vivo lentiviral vector development. Rather than treating each program as a standalone manufacturing challenge, we rely on a platform-based LVV framework that standardizes where standardization adds value - while preserving flexibility where biology demands it.
By anchoring programs in repeatable upstream production performance, defining downstream purification strategies, and analytically mature control packages, we can converge earlier on fit-for-purpose potency concepts and impurity reduction strategies. Just as importantly, a platform approach allows learnings from one program to be transferred systematically to the next.
This decoupling of target-specific vector design from core manufacturing and analytics reduces development friction over time. It supports comparability, simplifies regulatory dialogue, and provides greater predictability as programs move from early clinical supply toward larger studies and, ultimately, commercial considerations.
From what we see across the field, the long-term success of in vivo gene therapy will depend less on isolated technical breakthroughs and more on the ability to reliably manufacture, characterize, and control complex vectors across the product lifecycle. Programs that treat manufacturing and CMC as integral elements of development - rather than downstream obligations - are better equipped to sustain translational momentum.
Based on our experience, platform-based LVV manufacturing offers a pragmatic path forward. It transforms in vivo gene therapy development from a sequence of bespoke challenges into a reproducible, scalable modality - capable of supporting both scientific innovation and the realities of clinical and regulatory execution.
Stephen Madaras is a Business Development Manager (CDMO) at Miltenyi Bioindustry. He is a biotechnology process engineer specializing in cell and gene therapy manufacturing, with experience spanning process commercialization, characterization, and control strategy development. His work focuses on translating technical CMC concepts into practical execution and enabling scalable advanced therapy platforms that support regulatory readiness and long-term commercial success.
