Vaccine Viewpoint

Michael Kowolenko of Wyeth talks about the convergence of vaccines and biotech

By Gil Y. Roth

I met Mike Kowolenko at the BIO conference in San Diego in 2008. We had an entertaining conversation about his new role at Wyeth (see his biographical note on page 80 for more information on his title and responsibilities), discussing the quirks and intricacies of biomanufacturing at a major pharma company. Devoted to making sure that operations personnel grow as problem-solvers and people, Mike remarked to me, “If people don’t go out and experience new things, then they get their 20 years of experience as one year, 20 times over.” We decided to revisit our conversation this autumn and discuss the importance of vaccines in Wyeth’s biotech strategy.  —GYR

Contract Pharma: Tell me about how vaccines have impacted Wyeth and its operations.

Michael Kowolenko: The vaccine business is very important for Wyeth. We see it as one of our biggest growth areas. Prevnar has done extremely well, and we have a new version, Prevnar-13, that will be filed in the months ahead.

We’re really good at the bacterial expression platforms for vaccines, so that puts us in the infectious disease sector, Prevnar being the biggest example of that. In fact, a lot of the vaccines we’re pursuing are for infectious diseases.

Overall, I think, what you’ll see with Wyeth is that we’ll be more of a biotech company than a pharma company in the next couple of years.

CP: We’re seeing that trend from a number of Wyeth’s peers, aren’t we? Bristol-Myers Squibb refers to itself as a next-generation biopharma, while Lilly has just offered to buy ImClone. Do you think the not-quite-behemoth scale of these companies — in contrast with much bigger competitors like Pfizer, GSK or Merck — enables this strategy?

MK: Absolutely. Because we don’t have the large overhead associated with a company like those examples you mentioned — think of what their daily burn rate is to keep themselves going — we’ve developed a system that allows us to use different platforms, thereby letting us use more of our facilities to work on more of our products.

The whole idea when we went into biotech and now vaccines was to develop facilities that would let us support multiple products and different platforms, letting us shift depending on needs. That freed up resources that we could allocate for either in-house R&D or external partnerships.

CP: How far out do you have to plan to incorporate that level of flexibility?

MK: As we’ve developed greater robustness in protein expression, and better understanding with higher titers and yields, what we’ve found is that the facilities that we built, based on the designs of a decade ago, were somewhat constrained. In retrofitting these facilities, we increased their flexibility.

By designing support systems around reactors based on their relative titers, we were able to develop different scenarios for more production. So we’re seeing more retrofitting of older facilities, while newer ones have this flexibility built into them.

What we’re talking about is the ability to handle a wide range of buffers, with regard to volume, and reactor sizes. So instead of having a seed-train that terminates at 15,000 liters or 20,000 liters, you can take any one of the reactors that feed up to that reactor size and pull material and begin purification.

Biographical Note Michael D. Kowolenko, Ph.D. is Senior Vice President, Biotech Operating Unit, Technical Operations and Product Supply (TO&PS), for Wyeth Pharmaceuticals. Dr. Kowolenko is responsible for the development, manufacture, and global supply for all biotech products.      He joined Wyeth in 2007 after a 20-year career with notable successes in achieving global productivity and compliance gains as well as cost reductions for both major pharma and biotech companies.

CP: Do single-use/disposable bioprocessing systems play into this model?

MK: Not too much. Single-use can come in handy for doing clinical campaigns or perhaps products that you run relatively infrequently. But if you become adept at turning your facility over, there’s really no advantage in using those technologies. So it becomes a matter of employing single-use technologies in appropriate places. There’s not a one-size-fits-all model for this. It’s nothing to have a reactor suite that can also handle bags. You just build it all in, so it’s all plug-and-play. You try to design your facility to handle a wide range of production scenarios.

I can’t stress enough the need for flexibility. Even when you have a major biotech product that needs large-scale production, you have to be careful that you don’t build a facility for it that becomes a one-trick pony. Because as you increase efficiency in production, you still need to keep the facility utilized. That’s part of why you retrofit.

CP: How difficult is retrofitting a bio-facility in terms of regulatory burden?

MK: It’s not that difficult. The burden is on the company to demonstrate that it can run the facility in a state of control and show adequate separation between products and containment. It’s something the agency is quite used to, as the industry has tried to execute in the past. it’s actually the industry that has issues with trying to implement it.

CP: Why is that?

MK: I think they tend to view the regulatory agencies as unpredictable. My experience has been the opposite; if you come to them with a good scientifically and technically based argument, you don’t have an issue.

CP: Is there also a resistance based on the “big new buildings are prettier and make a bigger splash than retrofitting an older building” mentality?

MK: That can come into it. And if a facility is too dated, we won’t retrofit it; we’ll build a Greenfield site. Grange Castle has become a multipurpose facility.

CP: Obviously it’s a case-by-case issue, but generally how long does it take to retrofit a bio-facility?

MK: About a year. The thing to keep in mind is, as titers increase, the number of reactors you need decreases. So how much capacity do you want lying around, and how many facilities do you want to operate? You have to figure out the most efficient way to operate.

CP: What’s the biggest challenge in improving production?

MK: The challenge that I’ve taken on is to turn vaccines into biotechnology products. Unlike the way biotechnology was 10 to 15 years ago, we have a better understanding of the processes and what the critical product attributes are.

The difficulty with vaccines is that the proof of principle is in the clinical trial, and a lot of times you don’t know what the appropriate attributes are that you’re making. But that shouldn’t preclude us from trying to understand the physicochemical nature of the vaccines that we’re producing.

The arguments in the past have been that vaccines are too complex to characterize, too unpredictable. Well, that’s what I heard when I got into the biotech field 20 years ago, and we were able to untangle that to a significant degree, and get a better understanding of our processes. I see no reason why we can’t go after vaccines in the same manner. We’ve dealt with heavily glycosylated proteins that are difficult to characterize, such as FactorVIII. What’s the difference between that and conjugating sugars onto a carrier protein?

We need to overcome the dogma with data. We need to show that we can get a better understanding. The hope is that this will lead to an easier regulatory pathway, but even if it doesn’t, you’re still left with a better understanding of your process and how to control it. Reliability and reproducibility come from understanding the process, and that’s why Quality by Design is so important.

CP: Which do you feel is the bigger contributor to this understanding: better instrumentation or better conceptual models?

MK: The tools are much better than they were, but conceptually, there are some of us who accept that we can do this, that it’s not so different from what we were doing with proteins 10 or 15 years ago. It’s marrying that conceptual leap to the technological advances.

Again, I find it interesting how the arguments against doing this with vaccines mirror the ones I heard about biotech. But we solved those problems, for the most part. Proteins and sugars are what we do.

CP: Is it easier to characterize vaccines that are in development, or those that are already commercially produced?

MK: To me, it’s pretty much six of one, half-dozen the other. You have to understand what you’re going to find, and be able to answer the questions that come up. Some companies may be hesitant to go in and do extensive evaluations — PAT and QbD — on existing products, because they’re not familiar with this data.

Well, my response to that is, you should be able to ask the question, and you should use good science to pose a hypothesis that’ll answer it.

CP: What data do you think they may find?

MK: Variability of the process. What we do is conjugate sugars to this protein. It’s like pegylation of proteins. The big concern with pegylation was: what was the distribution of PEG on that protein? How do we characterize that and what does it mean for the pharmcodynamics of the product? We know that we put X amount of the sugars on Y amount of the protein. We don’t know exactly what the distribution is, and how that relates to a good immune response. But that doesn’t preclude us from asking the question. We have a general profile with the molecules. You may never get the exact correlation, but you should at least understand what influences the events. And that goes back to ICH Q8, Q9, and Q10. If you set up a good design space, you should understand those variables in development, understand how the process influences the outcome.

CP: What’s more difficult: managing new technology or managing personnel?

MK: Getting personnel to understand that manufacturing can be a dynamic and scientifically-driven activity, not a button-pushing activity. I tell people, “If you came here to push the same button for 20 years, forget it. What you’re going to be doing is looking at new processes and understanding them, looking at new technologies and figuring out how to apply them.” I want people who are thinking all the time.

It makes for a great work environment and more robust processes. Continuous improvement is much easier that way.

CP: How do you get operations personnel in that mindset?

MK: Move ‘em around! Get ‘em in different roles and positions, different parts of operations. Put them into quality, into proteins, into vaccines, so they get a wide breadth of exposure. What I’ve found over the years and the companies where I’ve worked is, if people don’t get opportunities to move into multiple areas and positions, they are more likely not to have the scope and breadth to ask the right questions and look at challenges from various viewpoints.

CP: How’s that approach working out for you?

MK: I think it was pretty successful in my “previous life,” and I think we’re getting there here at Wyeth. I’m only celebrating my first anniversary here, so it’ll take time.

CP: What’s been the biggest change in that year?

MK: We’ve got sites that act as a team. Personnel see that they’re not alone; they have difficult jobs, but they can share approaches, discuss successes and failures, ask for help, and support each other, making the whole team that much stronger.

CP: What’s the biggest area for improvement?

MK: We as an industry need to continue to embrace the newest technologies for inline monitoring of processes. We’ve implemented them to a certain extent, but we haven’t really pushed the envelope. We haven’t pushed PAT as far as we could.

CP: How difficult is implementing PAT in a biotech setting, as opposed to a small molecule one?

MK: The primary difference is that small molecules are much easier to analyze. The medium in which they’re usually isolated is usually an organic solvent. That’s a lot easier to deal with than one protein among hundreds of them. The complexity of the medium makes it difficult.

However, the technologies that we have now, with regard to sample processing, analytical columns for chromatography, electrophoresis, microsampling, mass spec and realtime analysis are all possible. We just have to use them to a greater extent.

We’re also using PAT with regards to gross measures of analysis in bioreactors. We’re measuring the different parameters: oxygen consumption, pH, and the like.

CP: Is PAT more useful for those gross measures, as proxies of what the final process will yield?

MK: If you do QbD right, and you understand the process, you can demonstrate control with it.

So if I have a process that, if it stays in the reactor for a certain amount of time, at a certain growth rate, I know that I’ll see certain glycosylation changes. I want to measure those and make sure they’re following the pattern that I saw when I was setting up my design. They remain constant as I move through the process.

We have to do the same thing for purification, developing algorithms for those analytics.

CP: How have you seen the industry change during your tenure?

MK: Today, we have to be even more cost conscious, and the best way to do that is to get people to think. Cost control is a challenge with vaccines, since government reimbursement is a major factor. The challenge is to make sure you can make it, get it globally distributed to everyone who needs it, and still make a profit for the company. One of our goals is to make these critical therapies products it available to all global patients who need them. If I can increase our output and get throughput and costs down, then I can get more products to more patients. That’s truthfully why I came here.

Gil Y. Roth has been the editor of Contract Pharma since its debut in 1999.