The Path To Producing An Insulin Biosimilar

By Tyler Menichiello, contributing editor

In 2020, the FDA redefined insulin as a biologic — a historic move that paved the way for the development of more affordable biosimilars. In its announcement, the FDA said it expects this regulatory transition to “enable a vibrant, competitive market,” one that empowers patients “by increasing choices and potentially lowering prices of safe, effective, high-quality medications.” One company traversing this regulatory pathway is rBIO, with the goal of reducing the cost of insulin with its biosimilar, R-biolin.

To learn more about what it takes to develop an insulin biosimilar, I met with rBIO co-founder and president, Cameron Owen, as well as its CSO, Dr. Deenadayalan Bakthavatsalam. They candidly told me about the company’s plan of execution, as well as how challenging it is to scale-up manufacturing to a commercial level (in the ballpark of several metric tons per year for insulin) — which is the company’s primary focus right now. Owen also walked me through the company’s genesis and the science at the heart of its operations — synthetically engineered E. coli.

The Path of Least Resistance

The FDA’s decision to forge a biosimilar regulatory pathway opened doors for companies like rBIO, but there’s another critical piece to the biosimilar puzzle. In 2013, the U.S. Supreme Court ruled that companies couldn’t patent naturally occurring genetic sequences. This means that the gene for naturally occurring insulin (also referred to as regular insulin, which is where the “R” in R-biolin comes from) is not patentable. “The reason we went after naturally occurring insulin is because there are no legal protections, we don’t have to fight against anybody else,” Owen says. “It’s a path of least resistance.”

That, and insulin is a large and growing market. It’s estimated that the global demand for insulin biosimilars is projected to increase at a 6.2% compound annual growth rate (CAGR). “We don’t have to guess and say, ‘If we build it, will they come?’ The answer is they’re already there, and they want us to build that bridge anyway,” Owen tells me.

But if wild-type insulin isn’t patentable, and anyone (with a bioreactor) can make it, then where is the commercial viability for a company like rBIO? That’s right — in the process.

“If you’re trying to play in a commoditized market — which most pharmaceuticals are — and you can produce higher yields per unit volume of your competitors, then you can always outcompete them on price,” says Owen. Of course, that means finding a novel way to produce and isolate pure insulin better than the competition, which is where rBIO comes in.

Cracking The Code With Synthetic Biology

“The whole premise of the company began as, ‘Well, let’s just create designer cells and see what we can do with them,’” Owen recalls. In 2020, after trial and error, the company had generated synthetic E. coli cell lines that were able to produce insulin, “albeit, at pretty small yields,” he says. “But that really became our avenue for commercialization: creating this IP to manufacture pharmaceuticals.”

This early work was optimized through a partnership with Dr. Sergej Djuranovic at Washington University in St. Louis. “From a 30,000-foot view, what Sergej and his group at Washington University found was a way to modulate protein translation efficiency to the point where it’s effectively constant,” Owen explains. This continuous over-expression of insulin forms the basis of the company’s IP.

Choosing The Right Cell Line: Ferraris Don’t Belong In The Woods

Almost as important as rBIO’s insulin is the company’s manufacturing workhorse — its expression cell line. When asked why the company chose to go with E. coli, Owen hit me with an analogy.

“If I were to say the word ‘car’ to you, out of the thousand different brands, styles, and models, one immediately pops into your head,” he says. “If I said, ‘I need this car to get me from point A to point B really fast on a racetrack,’ then a Ferrari is going to be your choice. But if I need you to get from point A to point B and you have to off-road and climb a mountain, a Ferrari isn’t suited for that, but a Jeep is.”

He continues. “To apply that analogy to science: If you were to make monoclonal antibodies, you wouldn’t want to use an E. coli system. You would want to use a mammalian cell system, because it more closely matches what you’re trying to do. It just so happens that the E. coli system is exactly what we need for what we’re trying to get it to do.”

It helps that E. coli is one of the “easiest and cheaper” systems compared to others, Bakthavatsalam adds.

“The technology just works really, really well in a prokaryotic cell line,” Owen tells me. “I mean, truthfully, before we started, I assumed that this technology would’ve worked better in the eukaryotic cell lines that we were testing, but the data and results came back opposite. Sometimes, that’s just science for you.”

Demonstrating Comparability In The Clinic

One benefit of developing a biosimilar is that the clinical trial is much less intense than it would be for a novel biologic. After all, it’s comparing the product to something that’s already been approved. After some preclinical work to show safety, “you can jump right into a Phase 3 clinical trial, which is basically comparability,” Owen says.

This involves enrolling patients who are already taking the reference product (in this case, Novolin R) and “switching them back and forth” between the reference and the biosimilar several times over the course of a month to show there’s no fluctuation or difference in how they manage their condition.

Before entering the clinic, rBIO had to complete analytical characterizations of R-biolin, which it did back in February. There are three main factors or readouts for a molecule to be accepted as an identical, functionally active biosimilar, Bakthavatsalam tells me. One is protein-protein binding — i.e., does the biosimilar insulin (in this case) bind to the insulin receptor on cells? The second is, if the molecule binds successfully to the receptor, does it result in autophosphorylation of the receptor? The third readout, and perhaps the most significant, is: does this autophosphorylation spur the desired sequence of events inside the cell? For insulin, this question is simple — is there uptake of glucose into the cells?

“These three functional attributes of insulin need to be addressed, and that’s what the analytical characterization has consisted of,” says Bakthavatsalam.

Scaling Operations And Finding The Right CDMO

Right now, rBIO is working to optimize its manufacturing process for a commercial scale. And, as any biotech knows, it’s not as simple as just doubling the recipe.

“When you’re scaling up — from 10 milliliters to 10 liters, or from 200 liters to 1,000 liters — at every step, there are nuances that have to be kept in mind,” Bakthavatsalam tells me. “Although your starting materials are the same, in due course, the cells behave differently. So, that’s why GMP involves trying to keep all the conditions ideal for the cell to produce the right product that can be extracted in the right way, consistently. That is the most critical piece.”

“I sometimes use the analogy of an internal combustion engine,” Owen says. In this analogy, the first step is akin to a go-kart engine. In the second stage, you need to build a car engine. “Even though the fundamentals of internal combustion apply, you don’t just take a go-kart engine and make the parts bigger, expecting it to work,” Owen explains. The process needs to be tweaked and refined for the scale you’re trying to achieve.

“We’ve got the car engine, but now, we need to go to a jet engine,” he continues. Again, the same principles of internal combustion apply, but it’s a fundamentally different engine. “You have to design each of these steps specific to the scale at which you’re working in,” Owen says. “From a biological standpoint, an insulin molecule has a 3D conformational shape with electrical charges and all sorts of characteristics — characteristics that, as you scale on each of these things, change how you need to interact with and treat it.”

A big part of ramping up commercial scale manufacturing is finding the capacity to do so — which, more often than not, involves outsourcing. The company decided to partner with a CDMO because it’s cost-effective, Bakthavatsalam tells me plainly. “To put together that kind of facility and scale, you need to go through a whole lot of regulatory processes to get approval,” he explains, “and then you have operational costs. That’s why we’re partnering with a CDMO. They have an existing facility and expertise in running those operations smoothly and in the compliance with the FDA.”

For rBIO, it was essential to partner with a CDMO that has experience manufacturing insulin. “Everybody cannot be an expert in every molecule,” Bakthavatsalam says. “We are focused on insulin, and we want the people who have walked that path. It makes our life easier to go through the whole regulatory process.”

Another reason rBIO had for partnering with a CDMO is to avoid what Bakthavatsalam calls “self-bias” — i.e., misplaced confidence in the process you’ve created. Like any good science, the process becomes validated the more it’s reproduced. “I truly believe a product can be commercially taken to market if it can go through multiple different layers and hands and remain reproducible,” he says.

As rBIO continues to scale and move R-biolin towards FDA approval, it’s my hope that more companies follow suit in bringing interchangeable biosimilars to market. Not only will a growing biosimilars market breathe healthy competition into the biotech space, but it should ultimately bring down the cost of these expensive medicines for patients.