Blog | January 24, 2023

4 "Hot Takes" On Cell & Gene Therapy Manufacturing From Advanced Therapies Week 2023

Source: Cell & Gene Collaborative
ARW Edit Headshot 2

By Anna Rose Welch, Editorial & Community Director, Advancing RNA


There is truly no article more difficult to write than a takeaway article from a conference. But as I recently spent the past week meeting people and attending discussions during Phacilitate’s Advanced Therapies Week, I wanted to attempt to do “the impossible” (for me): Share four of the burning takeaways that jumped out at me most during the week. (For a quicker glimpse at the first two days of the conference, I’d refer you to my previous LinkedIn post summarizing a couple of takeaways — a few of which I will touch upon again in slightly more depth here.)

Now, the following points may not necessarily fit your personal definition of a “hot take.” Manufacturing can only be so dramatic. And, frankly, if we’re doing it right, it shouldn’t be dramatic at all. But each of these arguments about how we should be approaching our manufacturing paradigms today come equipped with nuances that add a smidgeon of drama to this industry which, to be fair, aims to have a dramatic impact on patient health. In fact, when one panelist was asked what they were most looking forward to in 2023 and beyond, they didn’t provide a concrete event so much as express excitement over the fact we are working on treating and/or curing diseases that have thus far evaded science. To do this successfully, we must ask questions we have never asked before.

So, in light of this beautiful sentiment, I offer a handful of assertions and, yes, a few additional questions for us to mull in the year(s) ahead.

The Process Is Not The Product

Since jumping into the cell and gene therapy space in early 2021, I have heard and read the phrase “the process is the product” more times than I can count. So, when I hear someone say the complete opposite, I sit up a bit taller in my chair. (And frankly, I can count the number of times I’ve heard someone disagree with this manufacturing “cliché” on one hand.)

Throughout the conference, multiple experts presented on the importance of establishing a process using a quality by design (QBD) approach. Employing a QBD approach will certainly help companies understand how to better control — or, as one speaker said, tune — their processes to promote a higher quality product. This is especially true for those on the autologous side that are wrestling with variable patient starting materials.

But as we were reminded, we cannot forget that QBD was designed for highly characterized biological products, which our ATMPs currently are not. Not only do we understand biologics’ structure thanks to years of analytical development — and, in some cases, biosimilar development — but we also have scores of clinical data from which we can correlate our analytical understanding of CQAs with clinical effect.

On the ATMP side, we’re still trying to figure out how and to what extent shifts in the structure of our products will impact the clinical. Our current struggles with comparability are a fantastic illustration of just how challenged we are by the concept of structure as function. Our immature understanding of CQAs and their correlation to clinical effect were the two central reasons behind Kyverna Therapeutics CTO Karen Walker’s argument that the process is not the product — or at least it can’t be today for ATMPs. Instead, she argues that the product is the product.

As Walker explained: “Processes are built on modules that perform certain tasks. You can make sure your module is performing the task consistently and doing what it’s supposed to do. You can characterize the output from each of these steps to show that it remains in a state of control. But while the process is a mechanism to generate the product, you will not know if the product is ‘good’ or ‘right’ until you have understood it in the clinical setting.”

(If you’re interested in reading some additional debate and discussion on this topic, I’d refer you to the comments shared under my recent LinkedIn post.)

Volume Matters — But Don’t Sell Scale-Up/Scale-Out Short

During Virocell CBO Susan Nichols’ highly anticipated recap of the Top 10 events of 2022 in the CGT space, I was struck by two small details. In reference to Ferring Pharmaceutical’s (December) FDA approval for its adenovirus vector-based gene therapy in bladder cancer, Nichols explained that we won’t see the product on the market until the second half of this year as the company builds capacity to achieve commercial scale vector production. Similarly, our celebration of Janssen and Legend’s Carvykti approval for multiple myeloma can be tempered only slightly by the fact that the companies needed to employ a phased roll-out in a small number of treatment centers. This effort was certainly the responsible approach to ensure they could reliably meet demand — but it throws a spotlight onto the industry’s current scale-up challenges as companies enter the “real world” beyond clinical trials.

In this industry, we often refer to scale in terms of volume. Capacity and supply-chain constraints, variable starting materials, low viral vector titers, and high dosages are oft-referenced, volume-centric examples of why this industry remains constrained to smaller patient populations. I myself am guilty of limiting my definition of scale-up to just having “enough” of a product to meet commercial demand — be it a rare disease or a larger patient population. That’s why I loved this question posed to a panel on successful scale-up: “Is scaling just about volume?” The answers to this question revealed just how multi-faceted scale-up/scale-out is and should be.

At the most basic level, scale-up/scale-out is, yes, about achieving a volume that meets patient demand. But before you can obtain that volume, you need to know that the process you have in place — which encompasses the equipment, the raw materials, and the analytics (don’t overlook analytics!) — can also be scaled and that the product remains the same high quality at a greater volume. Afterall, a greater number/volume of cells is only valuable if they embody the same CQAs you need for your product. We also need our CDMOs, testing labs, and supplier partners to be able to keep up with our development needs, quality/supply requirements, and our patients’ demand.

All of these scale-up needs can similarly be applied to scale-out. However, as one expert argued, “scaling out” is perhaps less of a technical challenge than it is a logistical and operational challenge. Look no farther than this article outlining the types of expertise and how much staff a commercial cell therapy company will need to support a commercial process. Obtaining the necessary number of operators, training them (and keeping them trained), and organizing the movement of more people throughout facilities are scale-out challenges typically overshadowed by product volume constraints. Case in point: We don’t see industry headlines proclaiming, “Company X Fits More Employees Into Gowning Room To Achieve Scale-Out.”

We could also argue scale-out goes beyond the manufacturing facility. As one academic speaker shared, bringing the clinicians in early so they can interact with the product prior to being at the bedside is another aspect of operational scale-up we rarely discuss.  

We often compare our progress to that of the highly scalable large molecule/mAB industry. In turn, we tend to assume the only growth worth pursuing is that of making our processes/products “bigger” (i.e., more voluminous). But it behooves us to remember: We can only scale-up what we understand. Rushing through process development and analytical development puts us at risk of trying to scale processes that are not ready and/or may not be economically necessary for our patient populations in the end. And, as several conversations throughout the conference clearly demonstrated, there are a lot of pieces to scaling up/out that we need to be prepared to have in place beyond just producing greater volumes of product.  

We Are Not Currently Built To Be A Standardized Industry

Even I’ll admit that this statement is more “well, duh” than “hot take.” On the surface, we can say this based on the sheer number of cell types, viral vector serotypes (for AAV), or non-viral transduction or delivery vehicles with which we are working today. The fact that a good chunk of the CGT industry is autologous and is working with vastly different patient starting material is just another way in which we have had to get familiar with a lack of standardization. Some may argue — and, in fact, some did argue at this conference — that the sheer number of modalities with which we are working means the ATMP space(s) may never be standardized like the biologics space.

However, there are several additional characteristics of the general pharma/ATMP ecosystem worth noting that make pursuing standardization a bit more complicated. In the mAB space, for example, one of the critical steps toward standardizing in this space was the industry’s (particularly the CDMO industry’s) gradual alignment around standardized processes and unit operations. But as multiple speakers pointed out, establishing processes for one cell type will not necessarily carry over to a different cell type. Likewise, as the exhibition hall at Phacilitate clearly demonstrated, there is also a wide variety of technologies that all work differently despite being for the same unit operations. This variety makes aligning around a “singular”— or at least a more concentrated approach — near to impossible today. (Need I mention that comparability is expensive and hard, so exploring multiple technologies or having a back-up instrument for the same unit op are both quite difficult — especially on the autologous side of the industry.)

Outside of the variabilities between products and technologies, we also can’t overlook that this is a tricky economic climate in which to be a biotech. We’ve always been under pressure to be “first to market.” But now, we’re under pressure to convince investors that our science and our efforts to scale that science are the best and most deserving to garner limited funds. This raises a critical question in my mind: Will this extra competition for funding help or hinder the industry’s quest for standardization?

On the one hand, it could focus product development on those companies capable of demonstrating both clinical effect AND cost-effective operational/scientific scalability. As we move beyond the “drunken-sailor” age of biotech funding, investors are no longer keen on just funding ideas. In turn, this greater emphasis on data and more concentrated funding could (hopefully) ensure that we as individual companies are actively working on and are adequately funded to build the capabilities for what comes after R&D: CMC, QA/QC, engineering, scale-up/out, COGS, supply-chain maturity, manufacturing footprints, etc. Such efforts are critical for promoting greater standardization and more favorable COGS.

On the other hand, I could also see greater funding competition further fragmenting what is already a technologically fragmented industry. Should only a small handful of firms — each of which is focused on different cell types/modalities/technical approaches — garner funds, this could have a ripple effect on the space as a whole. In some cases, the argument could be made that the lack of more populous modality “ecosystems” could push us farther away from modality-specific standardization within the ATMP space.

[As I am not a psychic nor an investment guru, I’d love to hear any additional perspectives/thoughts/questions that are worth considering on this topic. Shoot me an email at]

Your R&D Team ≠ Aliens: Bring R&D Into The CMC Fold

R&D and manufacturing don’t necessarily have the same end goals in mind. That’s why I appreciated the brief — but still necessary — reminder that CMC professionals shouldn’t be afraid to work with and educate their R&D teams to foster greater alignment.

As former Spark Therapeutics CTO Cynthia Pussinen emphasized, CMC and tech ops teams will be the largest consumers of human and financial resources in the company. But with these great financial and human resource-based powers, we also have a great responsibility to provide insight into the ongoing “black box” of the work we do. This could mean more clearly explaining why we need certain (and/or so many) resources to achieve a quality, well-understood, and reproducible process. As Donna Rill, CTO of Triumvira, explained, even encouraging the use of higher quality reagents early on could be immensely helpful in easing the transition from R&D into process development.

As someone who spends their days communicating, it’s not surprising I’ll always be beating the drum for those on the CMC side of the industry to get more comfortable shining a light on what they do all day. And overall, as multiple panels/comments demonstrated, it’s up to us as CMC professionals to ensure our R&D colleagues understand what we ultimately need to achieve with our final drug products. Anything we can do to educate, collaborate with, and be better partners to R&D could help us more efficiently pave a road into development.