34 Viral Vector Manufacturing CDMOs

What is viral vector manufacturing and when do you need it?

Viral vector manufacturing is how you turn a gene therapy or cell therapy idea into actual drug substance. The vector, most often an adeno-associated virus (AAV), a lentivirus, or an adenovirus, is the vehicle that carries your therapeutic gene into a cell. Making it at quality and scale is a different discipline from designing the transgene, and it is the part almost every sponsor outsources to a specialist CDMO rather than building a suite in-house.

You typically need this work twice on the path to approval, and the two needs look different. The first is the GMP (or GMP-like) material for your IND-enabling toxicology and your first clinical batches: smaller scale, but it has to be representative of what you will eventually scale, because regulators will ask whether your tox lot and your clinical lot are comparable. The second is process scale-up as you move through the clinic, where the question shifts from can you make it to can you make enough of it, at a cost and purity that supports a commercial dose. Getting the platform decisions right early (suspension versus adherent, transient transfection versus a stable producer cell line) is what keeps the later scale-up from forcing a costly process change mid-program.

A few realities shape every viral vector engagement. AAV and lentivirus behave nothing alike downstream, so a site that is excellent at AAV may be a beginner at lenti. Yields are modest and analytics are hard: full-versus-empty capsid ratio, residual host-cell DNA and plasmid, replication-competent virus, and an accurate genome titer (usually by ddPCR or qPCR) are the specifications that decide whether a batch releases. And capacity is genuinely scarce, so lead times of many months to a clean GMP slot are common. These constraints are exactly why sponsors compare several qualified CDMOs rather than signing with the first one that answers.

What does a viral vector manufacturing CDMO actually do?

A viral vector CDMO owns the production process end to end, from the starting plasmids and cell bank through to a filled, released drug substance you can put into a tox study or a patient. The scope is broader than just running a bioreactor, and the strongest partners are explicit about which pieces they cover in-house versus subcontract.

In practice the work spans upstream, downstream, and analytics. Upstream is the production itself: a cell bank (often HEK293 or HEK293T for AAV and lenti), the transfection or infection step that drives vector production, and culture in either adherent systems (cell stacks, fixed-bed bioreactors) or suspension single-use bioreactors that scale better. Downstream is purification and concentration: clarification, nuclease treatment, affinity and ion-exchange chromatography, tangential flow filtration, and for AAV the full-empty separation that removes empty capsids carrying no genome. Analytics and quality control run alongside everything: identity, genome titer, infectious titer, potency, purity and impurity panels (host-cell protein, residual DNA, residual plasmid and helper), sterility and endotoxin, plus the safety testing for replication-competent virus (RCL for lentivirus, RCA for adenovirus). On top of that sits the GMP quality system, batch records, fill-finish coordination, and the CMC documentation your regulatory team needs for Module 3.

The other thing a good CDMO does is tell you honestly where your process sits. A research-grade transient transfection process that worked in a flask is not the same as a GMP, scalable, well-characterized process, and a candid partner will scope the development and characterization work needed to close that gap instead of quietly promising your bench process will just transfer.

How do you choose a viral vector manufacturing CDMO?

The decisive question is fit to your specific vector and stage, not the size of the name. A site with a deep AAV track record can be the wrong choice for a lentiviral CAR-T supply program, and a group set up for early Phase 1 material may not have the suites or the systems to carry you to commercial. Score two or three candidates against the same written scope (vector type, serotype or pseudotype, dose, number of batches, scale, and timeline) so you are comparing like for like rather than collecting quotes that measure different things.

Beyond the basics, the items below are where viral vector programs succeed or stall. Walk through each one with any shortlist before you sign.

• Quality and GxP status: confirm the suite and process are GMP for the material you need (IND-enabling, clinical, or commercial), and ask for recent FDA or EMA inspection history and any 483 observations. Non-GMP process development is fine and cheaper early, but your tox and clinical lots have to come from a system a regulator will accept.
• Capacity and lead time: viral vector slots are the scarce resource. Pin down a real GMP slot date and a milestone timeline (tech transfer, engineering run, GMP run, release), not a vague "we can fit you in," and ask what historically causes slippage.
• Modality and indication fit: match the CDMO to your exact vector (AAV serotype, lentivirus, adenovirus, or other) and platform (transient transfection versus producer cell line, suspension versus adherent). Their analytical methods and yields are vector-specific, so prior programs in your vector matter more than headline scale.
• Region and regulatory track record: check where the site is and which agencies it has supported, since you will file where the material is made and where you plan to sell. A clean record with the agency you are filing to is worth more than a marginally lower price.
• Data quality and analytics: ask how they measure genome titer (ddPCR versus qPCR), full-versus-empty capsid ratio, and potency, and how those methods will be qualified or validated. Inconsistent or poorly characterized analytics are a common reason batches fail to release or comparability falls apart at scale-up.
• IP and confidentiality: settle ownership of process improvements, any platform or producer-cell-line rights the CDMO claims, and how your construct, cell bank, and data are protected and transferred back to you. Treat ambiguous IP or cell-bank ownership language as a red flag, because the process is part of what you are paying to own.