Manufacturing or quality deficiencies represent 74 percent of the Complete Response Letters the US FDA issued between 2020 and 2024 (1,2,3). The key factor for success or failure is not clinical efficacy as biologics final users might imagine. In cell and gene therapy specifically, three high-profile programs were delayed or rejected in a single month in the summer of 2025 for CMC-related reasons alone. These are not isolated events; they represent a real difficulty in translating the quality strategy into a clear program with specific timelines and steps.

Looking at this, it would appear that the deficiencies showing up at the submission step trace back to decisions made, or deferred, months or even years earlier. The process is not failing at the end; it is designed without a clear logical path at the beginning. It can be fixed, but for that the conversation should start early in the program.

A regulatory expectation

The ICH (International Council for Harmonisation) quality guidelines are built around the manufacturing lifecycle with risk-based strategy applied from development through commercial phases. It is not a list of controls concentrated at the late stage of the process. FDA guidance on CMC (Chemistry, Manufacturing, and Control) for gene therapy INDs (Investigational New Drugs) has stated explicitly that early development decisions carry disproportionate downstream consequences (4). The agency’s January 2026 guidance on flexible CMC for cell and gene therapies reinforces the same logic (5): product quality controls, built with critical quality attributes, are expected to be in place early in development.

What this means practically is that a QC strategy cannot be redesigned at each stage of development. From bank qualification through GMP lot release, there should be one risk logic, with sampling parameters, acceptance criteria, and data packages connected to each other rather than rebuilt from scratch at every transition. I call this QC continuity. It is not a new concept. It is what the regulatory frameworks have been asking for, and what late-stage CMC failures suggest is still not consistently being delivered.

Innovation drives operational or inorganic growth 

Innovation in this industry is sometimes mistaken for an increasing list of methods, technologies, and platforms – as though access to a technique is the same as having the capability to apply it consistently, at GMP standard, exactly when a client needs it. It is not always that simple. The gap between a method existing and a method being reliably available inside a qualified quality system is where programs actually slow down.

Analytical methods that directly affect QC timelines, like RMM (Rapid Microbiological Methods) have been supported to access compendial status for cell-based products since EP 2.6.27 and USP〈1071〉(6,7). NAT (Nucleic Acid Testing) for viral detection has been part of regulatory expectations for years. NGS (Next-Generation Sequencing) was formally approved by ICH Q5A(R2) in November 2023 as a replacement for in vivo adventitious agent testing, with FDA and EMA adoption following in 2024 (8). QC sites must not only implement those innovations but also control, validate, and operate them, inside a GMP-compliant system.

From my view as a CEO, the strategic question is more critical than the technical one. You can license a method, but you cannot license the knowledge of how to apply it to an early-phase CGT product with limited material, or the validated workflow that connects it to a release decision, or the regulatory track record that makes the data package audit-ready. That knowledge is built over time and experience, or it can be acquired.

My approach has been carefully planned on this point. The acquisitions I would be inclined to move forward with in our group would not be designed to only add volume. They would be designed to target capabilities we do not have, or to establish local presence in territories where physical proximity to the sponsor and understanding of the local regulatory environment matter. Although it may be functioning well, a biosafety testing CRO (Contract Research Organization) that operates at a distance from its client’s manufacturing site, or from the regulatory authority reviewing the file, is perceived as a different proposition from one that is implanted in that ecosystem. Geography, in this industry, is not yet just a logistics question.

The implication for customers is direct. When they evaluate a testing or development partner, the relevant question is not what they list on their capability page. It is how they built those capabilities, whether organically or through acquisition, and what that means for the depth and consistency of execution they can rely on across the CRO program.

Constraints to take into consideration

Failure almost always appears in the same way. The QC program is built following a standard checklist, and it is later discovered that the plan does not match all the manufacturing needs. It results in add-on testing, rework, and time loss that could have been avoided.

One situation that illustrates this in cell and gene therapy is a QC strategy designed without taking the volume of material into consideration in the manufacturing process, leading to testing difficulties. FDA guidance on CMC for gene therapy (9) specifically acknowledges that compendial sample size requirements cannot always be achieved in this context and provides alternative approaches precisely for this reason (10). Another constraint to also take into consideration is the short timeline required for the development and manufacturing of those therapies. When the QC plan is not designed around those constraints from the start, the discovery could happen too late. By then, the program has already accumulated additional costs and release delays.

One operating model across regions

As programs increasingly combine development in one territory (for example: Europe, North America, or Asia) with GMP execution in another territory, the quality strategy needs to be designed to travel. For example, some regulatory frameworks like the EU and US share the same founding principles but differ in documentation formats, regional expectations, and specific method requirements. In our experience, when quality strategy differs between different regions, it is not due to divergence in terms of the science, but more from different operational habits at different sites that diverge during method transfer and create documentation inconsistencies and comparability gaps.

Again, if planned ahead during strategy development, geography could remain a logistics question and not become a QC hurdle. The same risk logic, the same method lifecycle discipline, and the same data integrity standards should apply regardless of which site performs the work. For programs running across regions, ensuring that expectation is set and tested early is consistently what separates a clean method transfer from a renegotiation.

Conclusion

To protect cost and timeline of the manufacturing process, manufacturers should consider three main points.

The first is to identify critical QC decisions impacting release before writing the QC plan. Building the QC plan around that list rather than around a generic checklist changes the resource allocation, the sampling design, and the documentation architecture from the very beginning.

The second is to check that sampling design matches what manufacturing can realistically provide at each development stage. In cell and gene therapy especially, where material is limited and each run has critical clinical outcomes, this question needs an answer before the QC plan is written, not after.

The third is to ask any manufacturing or testing partner directly how they manage method transfer, change control, and comparability data across sites and regions. That conversation will highlight operational risk faster than reviewing a capability list.

A lot of data and tools are available to design a seamless QC program early, such as: the regulatory record on CMC failures, guidance on what early-stage quality strategy should look like, compendial or recognized methods that reduce time-to-result and allow early detection. It is important to choose partners who have already made that type of decision themselves. Adopting the right strategy is not technically complicated, it is a matter of communication and forecasting and it is one of the key elements in drug development that is fully in the sponsor’s hands. 

Newsletters

Receive the latest pharmaceutical news, personalities, education, and career development – weekly to your inbox.

I have read and understand the Privacy Notice*

Subscribe