The BPI Theater is a 50-seat venue that for seven years has been located at the heart of the BioProcess Zone on the exhibition floor of the BIO International Convention. There, BPI provides attendees with four days of live presentations focusing on the latest scientific advances and business trends in biotherapeutic development and manufacturing. On Monday afternoon, 22 April 2013, Patricia Seymour of BioProcess Technology Consultants (BPTC) moderated a roundtable discussion on biosimilar development in the BPI Theater at BIO 2013 in Chicago, IL.
For a packed audience, she brought together four experts to provide valuable perspective: Uwe Bücheler of Boehringer Ingelheim, Richard Brudnick of Biogen Idec, Michiel Ultee of Laureate Biopharma, and Thomas Vanden Boom of Hospira. The “Introducing” boxes provide more information about each participant. Seymour’s own expertise covers development and commercialization of both biologics and small molecules, including supply chain strategy and operations leadership. Before joining BioProcess Technology Consultants, she worked for Millennium Pharmaceuticals, Covance Biotechnology Services (Fujifilm Diosynth Biotechnologies), ImmunoGen, the Dana-Farber Institute, and the Memorial Sloan-Kettering Cancer Center.
Seymour: Let’s focus first on some collaboration issues. Both Biogen Idec and Hospira have substantial collaborations (with Samsung and Celltrion, respectively). Richard, maybe you could go into more detail about how the synergy happened. How did you determine that the collaboration would be fruitful for both parties, what assets were brought to it, and how do you think it will transfer into real value creation for both parties?
Brudnick: At first you wouldn’t imagine that a Korean electronics company and a Boston-based biotech would come together as partners to develop biosimilars. It was quite a surprise to find that we share so many commonalities in outlook, in business culture, and in approach. The working relationship has been fantastic. We put the entire deal together in only about three or four months.
From a business perspective, it meets the needs of both parties. We were prepared to put in a little bit of capital, but only so much. We were prepared to offer manufacturing capacity and a certain amount of our P&L (profit and loss statement), but only to a point. And Samsung had the capabilities and willingness to take what we could provide, add to it, do more, and work with us where we could add to that as well. [The joint venture] Bioepis will move forward to a point at which we have an option to become a coequal partner. If it doesn’t fit our strategy at that point, then the company will continue on its way, and we will continue to be a licensor of products to it.
It’s interesting that we can work across a 13- or 14-hour time difference. Meetings tend to take place early or late in the day, but once people form working relationships, it’s amazing how effective they can work across the divide of hours and miles. So we’re very pleased with the progress of the joint venture so far, and I think it illustrates the commonality of training and insight and the underlying science that has become global.
As to what happens next, I think we’re all waiting to see how markets will unfold, what will happen with regulatory regimes, how these products will be differentiated on the market, and how much room there will be for how many biosimilars in each market.
Seymour: That’s the million-dollar question: What will this whole market look like in the foreseeable future? The story continues to unfold in Europe, and in the United States it’s still being written. But in some emerging nations, companies and products are already being established, and I want to talk specifically about that. But first, maybe Tom can talk about Hospira’s approach to partnerships.
Vanden Boom: I think there are two key business drivers around partnering: One involves gaining access to the full range of development and manufacturing capabilities. The other is risk sharing.
There is significant risk in this business. The audience here is familiar with development costs. In general, a single biosimilar molecule in the near term is likely to take US$100–200 million to develop, mostly depending on its therapeutic indication and the necessary scope of clinical trials. That’s a significant investment. To maximize a potential portfolio and identify partners that can share some of that risk makes sense.
Hospira is historically a small-molecule, specialty generic injectables company — not a biologic company. So in terms of quickly getting into this space, it made a lot of sense to partner. We’ve put together a global network of contract service providers and manufacturing partners to support our growing portfolio of products. For example, a notable partner early on was Human Genome Sciences, which is now part of GlaxoSmithKline (GSK).
OutsourcingSeymour: Mike and Uwe, when biosimilar developers looking for contract manufacturing capabilities come to your organizations, how do you approach those relationships? What are biosimilar developers looking for, and how are you positioning yourselves to be attractive service providers in that space?
Bücheler: Boehringer Ingelheim (BI) has been in the contract manufacturing space for 20 years, so it has valuable, trusted partnerships with many pharmaceutical companies and makes a number of originator molecules for them. Reliability of supply and high quality and compliance standards are the basis of our daily work. Biosimilar companies approach contract manufacturers to access that kind of expertise.
Introducing Richard Brudnick (Biogen Idec)
We asked our panelists to describe their experience with biosimilars, both personally and from the perspective of their companies. Richard Brudnick is vice president of business development for Biogen Idec, a biopharmaceutical company based in Cambridge, MA. Hear his introduction in the recording at the BPI Digital Media site online (www.bpidm.com) or right here
in the BPI mobile-app version of this supplement.
For BI, it was always a basic principle to have information protection procedures in place. We make sure that confidentiality is kept for each customer and for our own products. Being in the business so long with a sustainable contract manufacturing business should be an indicator that we can maintain confidentiality.
When companies approach us with their biosimilar projects, most have developed processes in their own R&D units or in collaboration with other companies. Often they come with a process and ask BI to transfer that into its operations. For us, that is straightforward because it’s what we do for originator molecules as well. We receive a process, we have a fast-track transfer or scale-up, and then we supply the product, typically for late clinical trials. With biosimilars, some companies approach early clinical trials with a final scale
and final process to minimize their CMC (chemistry, manufacturing, and controls) risk.
The more sensitive arena is process science and process development. We do develop biosimilars for our own business unit and supply product for those units. So for any company that approaches BI for contract development, delineation of activities for our own biosimilars has to be discussed and made very clear. But most companies approach with processes that have already been developed at other locations.
Ultee: Laureate has also been a CMO for a long time and learned about the importance of good customer relationships. One important aspect is to have a very robust project-management team. Even though we have just 120 total employees, five of them are full-time project managers. They are really the interface for everything that goes on between our scientists and those of a client company. One thing we’ve learned is the importance of good scientific collaboration so we can fully understand a given product. At the beginning, we have a scientific kick-off meeting where the team gathers to go through every step of a planned project — what we need to do and when — so there are no misunderstandings. It’s about the scope and what we’re going to propose and how we’re going to do it. We have a very open and transparent relationship.
We’ve worked on a lot of different monoclonal antibodies (MAbs), more than 40 of them, and those are some of the most prevalent biosimilar drugs in development now. We know a lot about the molecular structures and how to produce MAbs and look at the finer points of them. That’s really what biosimilars are all about: looking at the details. You need robust analytics to see the things that really matter in comparing your molecule with a reference or innovator drug. So our long track record of working with these molecules has been one reason people come to us for biosimilar projects.
Proving BiosimilaritySeymour: The next question is about demonstrating biosimilarity. The key is to have most CMC worked out before you go into the clinic. In innovator product development, things can change as you continuously collect clinical data. But with biosimilars, you don’t have that type of flexibility. So some companies are putting all their resources into CMC up front before their products go into a single patient.
One challenge for those companies involves products that were originally developed with old technology. Today’s technology (e.g., new expression technology) is much more efficient, much more productive, and probably would make a better product. Using aptamer technology, for example, you can show how heterogeneous an originator product actually may be. But if you start adapting and implementing modern technology for a biosimilar, how similar will the results be?
Ultee: Let’s start with a customer-developed process. We need to sit down and evaluate that process to see just how suitable it will be for scaling up. In many cases, it’s not really suitable. A customer comes to us for our capacity and experience in manufacturing. Often, discovery-based companies are not fully aware of the challenges in making a large protein molecule — especially a complex glycoprotein — in GMP (good manufacturing practice) manufacturing. So we look at the manufacturability and robustness of a process, we talk about the risks of taking a client that’s not quite ready, and guide clients toward something that’s less risky. And we try to come up with a consensus approach, with the understanding that we’ll do pilot engineering runs to prove a process at scale. For a biosimilar, we’re recommending several of those runs because of the importance of getting that consistent process. It’s very different from work with a phase 1 innovator product, for which a single run might be sufficient.
Seymour: Are you going to continue to match the heterogeneity of an existing originator product rather than making something that’s a little bit more homogeneous (and possibly a better product)?
Ultee: That’s an interesting question. Merck has presented on the GlycoFi system, showing that they tried to make epoetin with that. They had a beautiful single peak on the carbohydrate analysis, but that didn’t look anything like the original process. So it could not be considered biosimilar. The analytical chemists thought it was beautiful, but that really was not what’s considered suitable in terms of biosimilarity.
Introducing Thomas Vanden Boom (Hospira)
We asked our panelists to describe their experience with biosimilars, both personally and from the perspective of their companies. Thomas Vanden Boom heads the biosimilar R&D and manufacturing operations organizations for Hospira, a healthcare company based in Lake Forest, IL. Hear his introduction in the recording at the BPI Digital Media site online (www.bpidm.com) or right here
in the BPI mobile-app version of this supplement.
There’s a range to work within for all of these critical quality attributes. You look at different lots of an innovator product, and there will be some scatter in that range. You have to see just how much scatter. It will vary even among products sold in different regions of the world and made in different plants. So you want to see the total operating range that you have to work within and try to get as close to the center of that range as possible. That begins with your cell line because cells have a lot to do with how that pattern looks, with glycosylation and other posttranslational modifications accounting for a great deal of heterogeneity. The bioreactor process itself contributes as well, as do medium feeds.
You must be constantly comparing your product with the innovator drug. You’re trying to make it look as close as possible to the original, but how close? How similar? That’s the question that has yet to be answered.
Bücheler: Yes, I think the fundamental difference in biosimilar process development is determining that target profile, which is fundamentally different from developing a brand new process and profile for a new biological entity. Understanding your manufacturing process and the parameters you have to adjust to get your product into that target profile is a very challenging scientific task using technology that is 10–20 years more advanced. You have to be able to run your process in regulatory compliance, which is an additional challenge because you cannot just move back to old equipment. In many cases, the controls set up in such equipment are not suitable for today’s regulatory requirements.
This will not be an extremely crowded field; only a few companies will be able to overcome the challenges and meet those target profiles. That is the starting point from which you enter into your clinical investment. If you don’t meet that target profile, you probably shouldn’t spend the money only to find out that your product is not similar enough to the originator. Reference data will not be available. Available materials might change over time to a certain extent based on process changes and raw materials that are no longer available over a long period of supply.
Brudnick: In some ways, I think, developing a new molecule is easier than developing a biosimilar. In the old days, a biologic was defined not by what it was, but by how it was made. Then we had a revolution with well-characterized biologics no longer defined by how they were made, but by what they are. You could make it in different places with different processes, and it still was the same molecule. Now we’re finding that how it’s made st
ill matters.
Anyone who wants to develop a biosimilar will source innovator material from different manufacturing facilities and discover that there are differences among those. And our ability to characterize these molecules today is far greater than when they were first approved. So we’re measuring things that were never measured or standardized before and having to decide what’s important (to similarity) and what isn’t. The closer we get to that nonstandard target, the less expensive and less risky our clinical development activities will be.
When you’re starting out, you have to consider all of that in the context of what you can actually scale up from bench-top to large scale and still have a molecule that still looks like what you started with. It’s an incredibly difficult undertaking. And if you don’t invest the time and effort early on, you’re probably taking on enormous risk later. Most people didn’t appreciate that three or four years ago, when everyone was crowding into the biosimilar space and considering it an easy way to get into high-priced, more profitable therapeutics.
Vanden Boom: There are two keys. First is bioanalytics, particularly functional bioanalytics (bioassays). It really isn’t a question of whether or not there are differences, but rather whether or not those differences are clinically significant. That starts in vitro with an interrogation of the molecule using a full panel of functional bioassays. This area is driving a lot of innovative bioanalytical development efforts; biosimilar companies and service providers recognize that ultimately those advances may offer some relief from clinical trial requirements, which could have a huge financial impact. So the power of bioanalytics is key.
Another key is surveillance of originator products over a meaningful period (years). I’ll give a nod here to my colleagues at Sandoz/Novartis. You may have seen their paper documenting the bioanalytical profiles of three biosimilar drug targets over a multiyear period, with step changes occurring in the marketplace — not just structural bioanalytical characteristics, but in the case of one molecule, functional characteristics as well. Why is that important? That target you’re shooting for is not just the 2013 profile of the reference drug, but potentially the profile from a longer period encompassing postapproval changes made by the originator company. That changing range of product quality attributes (PQAs) is associated with the originator’s clinical and postmarket experience. And that is extremely powerful product knowledge for a biosimilar company.
At Hospira, such surveillance is a multiyear effort to monitor what’s going on with the originator products. For some products, it may be fairly straightforward. You won’t have access to the originator drug substance or API (active pharmaceutical ingredient). For some products, that’s not particularly meaningful because the drug product is really made from a formulated bulk. With other products, a deconvolution or deformulation is necessary to get to that window of observation of PQAs. It’s absolutely critical because in the end it will give you higher technical confidence that what you’re going after (the range of your product and process) is actually consistent with the decades-long marketing experience of the originator company.
Introducing Uwe Bücheler (Boehringer Ingelheim)
We asked our panelists to describe their experience with biosimilars, both personally and from the perspective of their companies. Uwe Bücheler is senior vice president of global biopharma operations at Boehringer Ingelheim, a contract manufacturing organization and research-based pharmaceutical company in Germany. Hear his introduction in the recording at the BPI Digital Media site online (www.bpidm.com) or right here in
the BPI mobile-app version of this supplement.
And the surveillance I’m talking about is not just routine release assays, but also high-powered characterization assays. Hospira and its counterparts are generating significant originator characterization data that eventually will find its way into biosimilar drug applications in the United States and in Europe. That new window of observation — not just into our products, but into the originator products — is something that health authorities currently don’t have.
Market OpportunitySeymour: Where are the market opportunities? We’re past the honeymoon of the first approval. Things have kind of cooled off in Europe. Several products are under review right now, but the most recent approval was a few years ago. Has the air gone out of this balloon? Where can we convert this enormous effort into actual commercial value?
Brudnick: Innovator companies will defend their products very effectively. But there is a place for biosimilars because payers will recognize that competition can moderate the cost of therapy for patients. How much competition will there be room for, how many competitors, and how much profit will vary enormously by country and product category. One thing that probably will be incredibly important to all is that it will be much better to get your biosimilar to market first rather than fifth.
Ultee: It’s going to take a lot of patience and deep pockets to develop biosimilars. Clearly there’s more to the picture than was initially thought. And I think the initial approaches (by the US FDA especially) will be very conservative, insisting on really very close biosimilarity. But requirements should loosen up over the years. We’re already seeing a lot of biosimilars approved in less-regulated countries, and those could be highly effective therapeutics for many people.
When that’s balanced against patient needs, maybe that’s the right approach. You’ll start to see whether some differences in molecular structure are really clinically meaningful, and I think that’s really the key parameter. At this point, we don’t know. But until it all shakes out, it will take a lot of time and money to develop these products. We’re already seeing some companies putting on the brakes a bit: e.g., the Teva–Lonza relationship on a Rituxin biosimilar.
Bücheler: For BI’s biopharma business unit, the overarching principle is affordable medicine for patients. And in that sense we are supporting new product development. We are supporting our strategic contract manufacturing partners. And we are also entering into the biosimilars field. It’s clear that there will be competition, which is good for patients and the affordability of drugs. I won’t speculate about pricing, but I think that the overall concept behind biosimilars is to provide high-quality products to more patients.
Vanden Boom: It is a dynamic space, and only companies that can navigate a very dynamic regulatory and competitive landscape will succeed. Even in Europe, where regulations have been in place for some time, the first monoclonal antibody is only now moving through that approval process. European regulators are gaining experience. The FDA regulations are still evolving, and I think that biosimilar companies can play a very important role in helping the agency frame what that regulatory science landscape will look like. That’s the approach that Hospira has taken.
The FDA recently communicated that it has held 38 initial meetings with biosimilar sponsors for 12 different molecules, which underscores the emerging competitive landscape. There’s a short list of biosimilar targets in the near term, and not all companies going after them will succeed. Those that do will internally and/or through partnersh
ips bring a very high level of science to the table. That will be required at least for regulated markets to gain credibility with health authorities and to bring the necessary data packages forth to support biosimilarity.
Seymour: We’ve heard some discussion of marketing an antibody (for example) in a less-regulated country first — just to gain some revenue and clinical experience — while continuing CMC development and so on to gather the information for filing in a western market (Europe or the United States). Tom, I’d like to hear your perspective first on taking that sort of bifurcated strategy.
Vanden Boom: In some of those markets, bioalternatives are currently acceptable. India is a good example. But a biosimilar is developed and eventually approved against a well-defined reference product in a well-regulated market. In less-regulated markets, development programs with minimal clinical trial requirements allow companies to get bioalternative products on the market. Those copy-cat versions would not meet European or US regulatory standards against a reference product. It’s a pretty tough competitive landscape, too, considering some of those local developers and manufacturers. So I think it’s likely to vary a lot by market or emerging market. Hospira looks at that case by case based on business dynamics.
Brudnick: As a matter of principle, I don’t think we would consider selling a lesser molecule into a lesser market with plans to sell a better one in better markets. We would hope to bring to those markets products on which consumers and patients can rely, knowing that they are made to the standard of the biosimilars. So I just don’t see that as a strategy we would follow.
Bücheler: Modifying the profile of a product in development means questioning the applicability of clinical trials or preclinical studies in which you have invested. I think the straightforward path is to target your product profile and then decide on investing in clinical trials. Typically that’s done according to international standards.
Ultee: I agree. It would be challenging and expensive to develop two versions of a protein.
Questions from the AudienceAt this point in the discussion, Seymour opened up the floor to questions from the audience. Because this session was held in the midst of a busy exhibition hall, she had to repeat questions for the panel members. So her questions here are reiterations of issues posed by members of the BPI Theater audience.
Seymour: How will push-back from innovator companies affect the final FDA guidance? Unable to get interchangeability out of the national legislation, some have discovered a “work-around” at the state level. A few US states are enacting (or at least voting on) related legislation. That may not be best for patients, but it’s certainly a creative way to protect a franchise.
Brudnick: There’s a very active public debate among innovator manufacturers, the generic industry, and public-health advocates. What’s most important is that the agency gets it right scientifically. Anyone with information or viewpoints to share should do so. It’s the FDA’s responsibility to balance that and act with due diligence to make decisions so that the industry can move forward.
Vanden Boom: A significant number of biosimilar companies (Hospira included) have been strong advocates for high scientific standards. Our expectation is that same level of scientific rigor needs to also be applied to policy. We’re opposed to any measures (at the state or federal level) that aren’t scientifically justified but simply reflect efforts to provide barriers or hurdles to the entry of biosimilars to the market. In the end, patients would suffer because of cost and limited access to important drugs.
Bücheler: In Europe, biosimilars are also an area of political interest. Some emerging markets encourage companies to manufacture within the country to get appropriate market access. So the political arena will influence manufacturing strategies, but ultimately the science should be successful. There probably will be delay factors here and there, but this will become a reality.
Ultee: It’s going to come down to money. When you consider what a biopharmaceutical costs a patient — $40,000–50,000 a year or more — something has to bring that down. But if you’re making billions a year on your innovator product, you’re going to rigorously protect it. That’s understandable. Innovator companies are going after the state legislatures because those are easier to influence, controlling market access by saying, for example, that a physician (rather than a pharmacist) has to approve use of a biosimilar. I think the FDA is more immune to those kind of pressures than state legislatures are.
Vanden Boom: When you have a product generating $26 million in sales every day, and you can put up barriers to biosimilar entry to prevent a biosimilar approval for even a week, the return is significant. Originator companies are likely to pull out all the stops to protect those multibillion-dollar franchises. It comes back to the science and the need for the biosimilar industry to advocate for scientifically based regulations and policies, so that regulations aren’t gamed simply to protect those franchises.
Seymour: Particularly the state legislations seem to be designed as obstacles. They aren’t going to stop progress on interchangeability, but rather slow down the uptake of biosimilars. You can still get a biosimilar approved, and doctors can still prescribe it, but it’s not going to be as easy to infiltrate the market. Once these biosimilars start gaining acceptance and prescribers get used to prescribing them, I think there will be enormous pressure from patient groups, advocates, trade associations, and so forth to remove that obstacle for interchangeability.
Brudnick: When the science allows for it, there should be interchangeability; before it does, there should not.
Editor’s Note: An audience member then asked more specifically about costs and pricing, especially in relation to the developing world.
Brudnick: We can look at the small-molecule generic industry to see what happens when barriers to entry are low, risk is low, and many competing products are not differentiated from one another. You would expect prices to reflect that, which is exactly what happens. We’ll have to see how the biosimilar market evolves. Barriers to entry are much higher. We’ll see how many competitors that allows and to what extent products are interchangeable, which may vary from one market to the next. Not everyone will be able to afford even a biosimilar, but more people may be able to afford these therapies than might otherwise have been the case.
Introducing Michiel Ultee (Laureate Pharma)
We asked our panelists to describe their experience with biosimilars, both personally and from the perspective of their companies. Michiel Ultee is chief scientific officer of Laureate Pharma, a contract manufacturing organization located in Princeton, NJ. Hear his introduction in the recording at the BPI Digital Media site online (www.bpidm.com) or right here
in the BPI mobile-app version of this supplement.
Bücheler: When we focus on developing a product to a target profile, not all processes will give the same high product yields and high product titers as for new biological entities. To meet your product profile, you may compromise on yield. That’s where large-scale facilities and operational excel
lence come in, where multiproduct facilities come in, because you probably cannot afford running a single-product biosimilar plant.
Seymour: What about developing a sensitive ADCC (antibody-dependent, cell-mediated cytotoxicity) assay so that you can monitor changes as you’re developing the biosimilar product?
Ultee: The ADCC assay is very sensitive to structures in the Fc portion of an antibody molecule, to minor changes in glycosylation. The plus–minus fucose difference, for example, is a very tough point to hit. And cell-based assays do have a high amount of variation, so they are a challenge in themselves. But you can perform high-powered glycosylation determinations and try to match an innovator drug, and you’ll see a range there too. But the final end point is your ADCC assay, which does unfortunately have that high variability.
Bücheler: Appropriately developed and precise assays are absolutely key. Ultimately when you present your data, a regulatory agency will go back to your assay and its capabilities. We have to accept that biological assays have some variability, so they will never have the precision of a mass-spectrometry instrument. An entirety of methodologies indicates whether or not you meet the similarity profile. Bioassays are key to that.
Vanden Boom: Bioanalytics — and in particular functional bioanalytic techniques — are absolutely critical to interrogating the function of a molecule.
Editor’s Note: Another audience member asked about intellectual property issues related to processes (rather than simply the products themselves).
Brudnick: Sometimes a process isn’t coming off patent when a product does. It’s often difficult to find some other way to make that product. And there is intellectual property out there that will extend well beyond the loss of patent protection for a molecule itself. You don’t want to face a legal challenge regarding infringement of a process patent. That’s a different issue than whether you can get FDA approval.
Bücheler: In biosimilar development, you don’t have access to the originator cell line. And you have no knowledge about the process that the originator uses beyond what is made public. So you independently develop a cell line and process into the corridor of your target profile. The patent landscape is very complex, addressing products along with certain processing steps and combinations. So I think your patent department is very important in process development so you don’t end up in a situation where you are not free to use your process.