Jie MA is a product marketing leader with over 12 years of experience in the CDMO and CRO industry. She currently serves as Director of Product Marketing at ProBio, leading global portfolio strategy and market intelligence for Biologics Discovery, Biologics CDMO and CGT CDMO business.
CREDIT: PROBIO
The promise of chimeric antigen receptor (CAR) T cells and other cell and gene therapies (CGT) has been limited by the hard reality that today’s treatments remain among the most expensive medicines ever developed. Complex, labor-intensive manufacturing workflows, long production timelines, and fragile logistics have all constrained scalability and limited patient access worldwide. While manufacturers are actively working to reduce costs, these challenges have so far placed a ceiling on how broadly conventional cell therapies can be deployed.
As a result, interest is rapidly shifting toward in vivo approaches. Rather than engineering cells or proteins in external bioreactors, in vivo therapies seek to turn the patient’s own body into the bioreactor, programming cells directly to produce therapeutic molecules. This shift has been enabled by the maturation of delivery technologies such as viral vectors and lipid nanoparticles (LNPs), the same class of systems used in COVID-19 vaccines. Since then, RNA-LNP platforms have continued to evolve, offering faster development timelines, modular design, and greater manufacturing flexibility compared with traditional biologics.
To explore how these scientific and manufacturing trends are reshaping expectations for advanced therapy development, DDN spoke with Jie Ma, Head of Product Marketing at ProBio, about the strategic decisions, platform technologies, and end-to-end integration required.
The CGT field is often described as entering a scale-up phase. From ProBio’s vantage point, what are the most significant shifts you’re seeing as programs move from early clinical development toward commercialization?
What we're observing is less a gradual evolution and more a fundamental rethinking of how in vivo CAR T therapies are built. By removing the complex ex vivo steps — cell extraction, engineering, and reinfusion — in vivo CAR T programs can be designed with commercial scale as a starting assumption rather than an afterthought.
In practical terms, this means end-to-end integration has become a prerequisite rather than a preference. Early-stage design decisions around plasmid architecture, formulation strategy, and analytical frameworks must now anticipate the demands of good manufacturing practices (GMP) from day one. Programs that treat these as later-phase concerns consistently encounter costly rework.
We're also seeing a decisive shift toward modular delivery infrastructure, particularly RNA-LNP and targeted lentiviral systems, where scalability, speed, and cost efficiency are engineered at the platform level. What distinguishes the programs advancing most confidently toward commercialization is not just the therapeutic concept, but the manufacturing architecture underpinning it.
With more than 160 Investigational New Drug (IND) approvals supported globally since 2017, what common manufacturing bottlenecks still catch developers off guard as they advance into later stages?
Despite the field's maturation, certain failure points recur with striking regularity. For both viral and non-viral delivery systems, the most consequential, and most underestimated, is plasmid DNA quality. Plasmid is the foundational input for both approaches, yet its design and qualification are frequently treated as a procurement exercise rather than a strategic chemistry, manufacturing, and controls (CMC) decision. We have seen programs delayed by six months or more because plasmid specifications established at the discovery stage were not CMC-ready.
For mRNA-targeted LNP, there are several distinct CMC challenges. First, the targeting antibodies are typically antibody fragments, which introduces inherent hurdles from low expression yields to complex purification workflows. Second, scaling these constructs from lab bench to clinical or commercial production is far from straightforward. Third, the antibody-to-LNP ratio has to be carefully optimized for each candidate, because it directly impacts both in vitro and in vivo efficacy. And fourth, the analytical methods in this space are still immature — there's no established playbook, so you end up doing extensive in-house development for every program.
Targeted lentiviral vectors (tLVVs) present an entirely different set of challenges. On the upstream side, engineering the VSV-G envelope (vesicular stomatitis virus G), a widely used viral surface protein, and adding plasmids to encode the targeting antibody can reduce crude titers and compromise overall transduction efficiency. Downstream, vector aggregation becomes a real concern once you modify the native envelope proteins. You also need more extensive characterization and stricter controls for impurities — host cell protein, for example — because the safety bar is higher. And finally, you can't rely on off-the-shelf analytics for the drug product; you have to develop tailored methods in-house. That means having a comprehensive and reliable in-house analytical platform isn't a nice-to-have — it's a prerequisite.
How are sponsor expectations changing when it comes to Contract Development and Manufacturing Organization (CDMO) partnerships — particularly around speed, flexibility, and platform readiness?
The transactional CDMO model, discrete services contracted in sequence, is giving way to something closer to a development alliance. Sponsors now expect a partner capable of holding the entire program thread, from CAR lead discovery through to GMP manufacturing, without requiring the developer to manage handoffs between disconnected service providers.
Speed remains critical, but it is now inseparable from platform readiness. Sponsors expect modular capabilities for RNA-LNP and viral vector systems that can compress IND timelines without sacrificing process robustness. A CDMO that can move quickly but lacks an established analytical platform for the chosen modality creates a different kind of risk.
Flexibility has also taken on a more specific meaning. Sponsors need customized CMC solutions tailored to each novel modality and regulatory pathway — because no two programs follow the same trajectory.
How do you see the balance evolving between viral vectors such as lentivirus or AAV and non-viral systems like targeted LNPs in in vivo CAR T development?
Rather than converging on a single dominant modality, we expect the field to settle into a complementary landscape where delivery-system selection is determined during project planning and guided by therapeutic objectives, target biology, patient population characteristics, and manufacturing considerations.
Targeted lentiviral vectors remain highly relevant where durable expression is required and the regulatory pathway benefits from established precedent. For applications where transient expression is therapeutically sufficient, or where the safety profile of a non-integrating system is advantageous, RNA-LNP systems offer compelling advantages in scalability and manufacturing flexibility.
Looking ahead over the next three to five years, we expect RNA-LNP to gain significant ground in autoimmune diseases where repeated dosing is clinically feasible, while targeted viral approaches will likely retain their position in hematological indications requiring stable, long-term CAR expression. The key inflection point will be clinical efficacy data from the first wave of in vivo CAR T programs currently in early trials.
The industry often talks about “platformization” for CGT manufacturing. Where do you believe true standardization is realistic today — and where customization remains unavoidable?
Standardization is genuinely achievable and already being realized in the foundational components of the workflow. Plasmid DNA production, antibody fragment development, in vitro transcribed (IVT)-mRNA production are areas where platform approaches deliver consistent quality and predictable timelines without compromising program-specific requirements. We’re also continuously accumulating experience to optimize tLVV and tLNP platform process.
Customization, however, remains essential at the therapeutic interface. CAR construct design, targeting antibody selection, and the optimization of delivery system for a specific indication are inherently application-specific. The biological complexity of in vivo CAR T, where the same delivery system must navigate different tissue environments and immune contexts, means that one-size-fits-all solutions are not yet realistic at the clinical level.
At ProBio, we see these as complementary rather than competing priorities. Platform efficiencies in upstream and analytical processes create the capacity and consistency needed to support deeper customization where it actually determines therapeutic outcome.
What advice would you give to developers entering the in vivo CAR T space for the first time, particularly when selecting partners for CMC and manufacturing support?
The most important principle is to select for genuine end-to-end capability, not an aggregation of isolated services. The handoffs between discovery, CMC, and GMP manufacturing are where programs lose time and coherence. A partner that can hold the scientific and operational thread across all three phases removes a category of risk that is easy to underestimate at the outset.
It is also important to recognize where the field currently stands. Most in vivo CAR T developers are focused on achieving the first major clinical breakthroughs and advancing programs to market as quickly as possible. At this stage, demonstrating safety, efficacy, and manufacturability typically takes priority over extensive optimization of cost of goods. As a result, partner selection should be driven first by technical capability, development speed, and regulatory execution rather than purely by manufacturing economics.
Beyond breadth of capability, prioritize multi-modality expertise. The in vivo CAR T field is evolving rapidly and the optimal delivery approach for a given program may shift as clinical data emerges. A partner with deep experience across both viral and non-viral systems provides strategic optionality, not just execution capacity.
Finally, evaluate the quality of scientific collaboration, not just the service catalogue. The programs that advance most efficiently are those where the developer and CDMO operate as genuine co-developers: sharing data early, stress-testing assumptions together, and adapting rapidly when results demand it. In a field moving this quickly, that kind of alignment is a key competitive advantage.
