The Monster Under The Bed: Defining Potency in the mRNA Space

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

Those of you who know me or have read past columns/watched past videos will likely know that, in addition to being a poet, I also grew up as a violinist. And, as it turns out, the violin is an instrument ripe with metaphorical opportunity in the world of pharma.   

I was reminded of this a few months ago when I found myself in the Musical Instruments Museum in Brussels. I was particularly struck by the number of physical evolutions string instruments went through to arrive at what we know as the modern-day violin, viola, cello, and string bass. As some of the pictures in my LinkedIn post will demonstrate, violin makers explored some pretty wild shapes in an effort to improve a violin’s ability to project and direct sound (i.e., their expression... see where I’m going with this?) In most cases, the shapes were either too unwieldly to handle/feasibly play, or the physical properties of the instrument did not result in the desired acoustic properties luthiers were seeking.

There are many ways I could spin these fun violin-related factoids to fit the world of mRNA development. But having recently had a conversation about the complicated trilogy that is the expression, function, and potency of our mRNA products with the mRNA industry’s own Khaled Yamout, it seemed most appropriate to think about what the violin “industry” did to improve the “expression” of its “products” as we work to understand how and why our mRNA expresses and functions the way it does in vivo.

During the CASSS mRNA Symposium last spring, a seemingly simple, straightforward question was raised: How are we defining and measuring our product’s potency? But as we already know, this is far from a straightforward question — and, dear reader, based on the resulting discussion amongst regulators and panel members, it became clear we have a lot of different definitions on what potency even is in the mRNA space.

Following this panel discussion, I sat down with Yamout to dig more deeply into this topic. At the surface level, Yamout defines mRNA as a bridge between a small molecule and a biologic. The mRNA is produced synthetically and used elicit a biological entity to achieve a biological activity. In turn, this is a distinction that requires “a new definition of potency for mRNA.” But as our conversation continued, an even more interesting argument emerged — one that suggests perhaps we haven’t gotten to know our molecules and their function(s) well enough to even begin to arrive at the appropriate definition of potency today.

Debating Expression Vs. Potency

This is not my first foray into writing about the beast that is potency — or, as I once termed it, the ATMP industry’s very own “monster under the bed.” I’d encourage you to read these two articles (here & here) to get up to speed on just how intensely and thoroughly this topic has been previously discussed/debated in the ATMP space. Not only do I dig into what makes demonstrating the potency of our products so difficult at this juncture, but I also discuss how we and regulators have been striving to define the most important step(s) of our products’ biological cascade. (I won’t be recreating that wheel here in this article for the sake of brevity.)

Now, to be clear, we’re not without guidance on potency assays. The FDA released its “Potency Assurance for Cellular & Gene Therapy Products” draft guidance back in December 2023. More recently, however, the mRNA space got its own definition from the MHRA in its draft guideline on individualized mRNA cancer immunotherapies.

As the MHRA guidance reads, “Potency is the quantitative measure of biological activity based on the attribute of the product linked to the relevant biological properties. The potency assay should be based on the intended biological effect and ideally related to the clinical response.”

Though there is, arguably, a lot to unpack there, what is perhaps most important to note about this section of the guidance is the MHRA’s specification that protein expression is not to be conflated with potency. In fact, as I’ve written before and as previous conversations in the ATMP space have emphasized, expression and function/biological effect are two related but separate steps in our products’ biological cascade in vivo. As the MHRA guidance specifies, “An analytical method that measures a correlate — such as protein expression — is considered a functionality assay rather than a true potency assay.”  In other words, expression ≠ potency.

Now, I will acknowledge that relying on expression to demonstrate clinical benefit/a product’s biological function (i.e., its potency) is certainly not without precedence. Back in 2023, I wrote a column about Sarepta’s gene therapy entitled “FDA to Sarepta’s Gene Therapy: ‘Express Yourself…Maybe?’” In this column, I discussed at length the tenuous relationship that exists between “expression” and biological function. As Sarepta argued, expression of the micro-dystrophin protein should be “enough” to predict a product’s clinical effect/benefit, and, at the time, FDA adcom reviewers remained dubious.

Likewise, back in December, the ATMP space celebrated the approval of the autologous cell therapy Waskyra, which was notable first and foremost because the BLA was filed by a nonprofit (Fondazione Telethon ETS). But what got surprisingly few (actually, no) comments was this noteworthy trinket from the FDA’s summary basis for regulatory action document: “FDA permitted a potency assurance strategy based on a [Wiskott-Aldrich Syndrome Protein/WASP] expression assay and an interim analysis demonstrating correlation of WASP expression values in the final product with increases in platelet count between baseline and Day 180, a relevant clinical outcome, rather than a drug product functional potency assay.” (Please note, the work isn’t done yet; they do have several post-marketing commitments they must meet for potency assurance.)

I bring all of this up because, in the mRNA industry, expression is “God.” Much like violin makers, our molecule design decisions — whether it be sequence and UTR design and modifications to the 5’ and 3’ ends — are made to achieve optimum rates of translation and levels of protein expression. We often boil down our likelihood of success (i.e., protein expression and the desired clinical effect) to three physical characteristics: Does our mRNA have a cap, a tail, and the “right” sequence? If it has these three things, it will express the necessary protein and we are golden. If it doesn’t, it will not express — and without expression, there will be no clinical effect. Such emphasis on having all the right components (i.e., caps and tails) has shaped us into an industry that strives to achieve the highest capping percentage, sequence integrity, and tail lengths.

Likewise, such seeming structural simplicity has also informed our approach to demonstrating potency. It’s not uncommon to hear arguments that mRNA is strictly the vehicle with which we inspire the body to produce the necessary protein. In turn, this lends itself to arguments that measuring potency can/should be done by ensuring the DP reaches the cell’s nucleus and simply expresses “x” amount of protein.

Why We Need To Take A Closer Look At Structure/Function First

However, as my conversation with Yamout went on to suggest, there still remains a lot to understand about the structure and function of our molecules — and, at this point in time, there’s still, arguably, much more to learn about the proteins that our mRNA is creating in vivo to successfully argue two things: One, that expression will be enough to demonstrate the potency of our products; and two, that certain structural goals/metrics we’ve established — such as different cap structures and tails — are truly necessary to achieve optimal clinical effect and to enhance the half-life of the mRNA in the cell.

“We need to be linking physical characterization to functional characterization in the cell,” he said. “We need to do all of our in vitro and in vivo studies on the protein itself and attain the data demonstrating that, based on my sequence design, I’m getting the right, properly folded protein and that I can produce the necessary quantity of protein per my mRNA-LNP dose. This will help us better understand how specific physical quality attributes values that are measured analytically correlate to functional data.”

In some cases, our general lack of understanding is reflected in our desire to achieve certain goals with our mRNA production — for example, a high percentage of capping efficiency. However, in Yamout’s experience, we rarely have the functional data to support these goals. This raises the question of whether we’re striving for heights for heights sake rather than for the benefit of our product’s function.

“Every time I talk to a company, I ask about their capping efficiency, for example,” Yamout offered. “Often times, that company will respond that it’s at 90 percent but that it needs to be higher than that. My question in response is, ‘Why? Where is your cell functional data to tell you that you need to achieve 95 percent capping efficiency?’ Very few have done these experiments.”  As he went on to explain, this kind of thinking raises other questions; for example, how is the protein expression impacted if you have a capping efficiency of 85 percent vs. 95 percent, and is that difference clinically significant? Not to mention, different cap structures do have an impact on how much protein is expressed.

Not surprisingly, there are a lot of other factors that influence the protein being created — be it the chosen cell line for testing, UTR design, codon optimization, and translation time. On the one hand, as we already know, each cell is its own kingdom with its own biological rules. Not all cells are created equal, and not all mRNAs are translated equally within a cell — and, to be clear, there is no standard cell line within which we’re all working. Likewise, the design of the 5’ and 3’ UTRs, as well as codon optimization, all influence the rate of translation, which, in turn, can impact the quality of the protein produced.

 “If you over optimize your codon, the rate of translation becomes incredibly fast,” Yamout added. “When that happens, you’re going to get a large amount of protein fast, and it might not be folded properly,” Yamout added. “If the protein is not folded properly, it’s going to impact the protein binding to the desired target. So, your product may have produced the required amount of protein, but you also must make sure that it’s binding properly to the antigen that you’re after.”

Aligning Around A New Definition of Potency

In past discussions around potency in the ATMP space, the regulatory agencies have emphasized the importance of understanding and demonstrating the function of the protein we’ve created in vivo. However, in the mRNA space, we (so far) seem to have a variety of different definitions of what the function of our mRNA products is. Does it simply mean the expression/translation of the right protein? Does it mean measuring the amount of protein being translated? Without a clear, unified definition of our products’ functionality, we’re unlikely to align around a clear(er) cut definition of potency.

Hence why Yamout points to our need, first, to better understand the functional characterization of our products, as well as why we need to align around a definition of potency for mRNA. In the small molecule world, potency can be determined by chemical purity and concentration. In the biologics world, the question we ask and answer is whether the protein is binding to the proper antigen.

If we follow in the footsteps of the biologics space — and guidance in the ATMP world — this would mean, for mRNA, we also need to test the functionality of the protein we create in vivo to demonstrate potency. However, as Yamout added, this will require a standard to properly move forward.

“If I’m making an mRNA product and I’m creating a specific protein, am I testing the functionality of the protein?” he asked. “If I’m going to call this potency, the question becomes ‘potency relative to what?’ If we’re going to define potency as the functionality of the protein, we are going to need to have a reference standard against which to test our translated proteins. So, if we are considering mRNA a biologic then we need to test for protein expression to determine the dose of the protein expressed and the biological activity of the expressed protein.”