More Than The Rise In Biologics Prompts Reorg At FDA
By Rob Wright, Chief Editor, Life Science Leader
Follow Me On Twitter @RfwrightLSL
Change is very, very difficult to do in government,” says Steven Kozlowski, M.D. In 2012, the director of the FDA’s Office of Biotechnology Products (OBP) learned that the Center for Drug Evaluation and Research (CDER) would soon undertake a major restructuring. CDER had decided to reorganize around drug quality manufacturing, potentially including biologics. The plan would really change the Office of Pharmaceutical Science into the Office of Pharmaceutical Quality, and it would focus on greater integration between review and inspection to achieve better manufacturing consistency. “The reorg would fulfill a lot of ideas that had been talked about for a long time, like GMP [good manufacturing practice] for the 21st century, quality by design [QbD], and a whole variety of advanced manufacturing topics,” explains Kozlowski. “But there were questions about where OBP, the office responsible for reviewing the manufacture of all biologics in CDER, would fit into the plan [i.e., would OBP be integrated into this new system or be treated differently?].” For example, biosimilars were on the rise, so it would be beneficial to keep a group capable of dealing with the anticipated surge of biosimilars submissions intact.
Eventually the decision was made to leave OBP pretty much untouched. You probably wouldn’t be surprised to learn that for many OBP employees the notion of being able to continue “business as usual” was welcome news — at least initially. “A lot of people don’t like change,” Kozlowski affirms. “But a number of people both inside and outside of OBP approached me saying, ‘If a lot is changing within CDER, by doing nothing are we losing an opportunity to participate in a change that will better prepare us to do our job?’” he states. “That question was the aha moment for me.”
Despite OBP not being part of CDER’s formal reorganizational plans, Dr. Kozlowski did what many government critics might consider unthinkable — he seized the opportunity to implement change when not mandated to do so.
IS THE BIOLOGIC SUBMISSION BUBBLE ABOUT TO BURST?
The decision to make a change had a lot to do with an analysis of current market trends for biologics. “In 2014, we had a little more than 170 original INDs [investigational new drugs],” Kozlowski shares. “Not all of those necessarily had full manufacturing review, which is our primary responsibility.” The OBP director estimates there are presently more than 1,300 active biologic INDs within the FDA. The number of biologics being approved isn’t nearly as large as that of pharmaceuticals — yet. “I think we had 11 approvals in 2014, but that number doubles every 10 years or so. In the 1980s, there were only about two a year. In 2000 it doubled to about four a year. Since 2010, it’s been about eight a year.” Considering this trend, combined with the Biologics Price Competition and Innovation Act (2009) moving some additional products into being regulated as biologics, Kozlowski expects another doubling — soon. “Predicting biologics growth is very tricky,” he admits. “I estimate it is actually going to grow faster than the current rate.” Why? He says there will be continued growth in novel products, the primary contributor to the biologics doubling trends, on top of which you will also have an increase in biosimilars and perhaps the continued rapid growth of antibodies. “If you look at antibodies as a subset of biologics, I think we have close to 60 marketed products,” he estimates. “Whether they are antibodies or antibody fusion proteins, there are a lot of INDs for them, and that subset may grow fast for a while, too.”
Predicting biologics growth is very tricky. I estimate it is actually going to grow faster than the current rate.
Steven Kozlowski, M.D.
Director of the FDA’s Offi ce of Biotechnology Products
The predominance of antibodies in the biologics space is one reason why OBP, prior to the recent reorganization, used to consist of just two divisions organized around product structure. “You know, one for antibodies and one for everything else,” Kozlowski laughs. This structure may have made sense when OBP was part of the Center for Biologics Evaluation and Research (CBER) or even in the early 2000s when it was first moved to CDER. “The growth of antibodies created some unevenness within OBP,” he says. “It’s hard to predict what’ll be hitting our desks over the next 5 to 10 years. Although we know antibodies are growing, we really don’t know by how much and what percentage of IND applications they will represent or even whether there’ll be some new thing to consider.”
Once the decision was made to reorganize, the first challenge facing Kozlowski and his team at OBP was what structure the new organization should take. “If you try to reorganize while trying to be prophetic about what will happen in the future, if your industry and marketplace predictions aren’t accurate, you will just end up needing to reorg again,” he explains. This was something he was hoping to avoid.
GROWING FROM TWO TO FOUR DIVISIONS
One of the first issues that had to be resolved was how the workload would be best managed.
“We were operating with 50 to 60 FTEs [full-time equivalents per division], which is a large number of people for someone to manage,” says Kozlowski. His initial idea was to just add another division, which he discussed with some of the managers involved and also with the managers in OPQ (the Office of Pharmaceutical Quality), under which OBP resides within CDER. The consensus was that just adding another division would be too shortsighted. “If we were at capacity when we started, we haven’t really achieved much,” he concedes. “That drove the idea of creating four divisions, with each being small enough to be managed effectively but still having the ability to grow a little.” As a result, four new divisions of biological product review and research (DBRR) were created and simply named DBRR 1, 2, 3, and 4.
The next step was to determine how to divide the work. “That process involved a massive number of pie charts,” Kozlowski shares. “We realized we needed other ways of dividing product classes.” In the past, for example, when the antibody load became too high for the antibody division, IND products that consisted of fusion proteins were shared with the other division. Kozlowski says they considered an option that included continuing to divide antibodies in this manner while also dividing enzymes separately from cytokines. In fact, they considered multiple options and configurations — a process Kozlowski describes as almost an exercise in futility. “You could divide this work in a variety of ways, each of which makes sense on its own right now, but you would likely be constantly tweaking the number of divisions depending upon what’s dominant in the marketplace in the future.” Ultimately, they decided on divisions that review therapeutic proteins, whether antibodies or not.
This idea of having four divisions within OBP that didn’t have specific areas of expertise got a lot of pushback. Kozlowski says some of the questions surrounding this decision included, “How would we assign work, then? How would we maintain expertise? How would we ensure consistency in four groups where everybody does everything?” In order for this new noncategorized division system to work, OBP would need some means to control assignments centrally.
AN ALGORITHM TO MATCH EXPERTISE TO PROJECTS
As part of the reorg at OBP, there existed a group of individuals acting as project managers. “These people were managing review assignments,” Kozlowski says. “Life cycle products tended to go to the group that previously reviewed them to ensure continuity.” This was fairly straightforward. However, when it came to assigning new work, the challenge was to make sure you had the right type of expertise for each assignment. “That led to the idea of what we’re calling ‘the algorithm.’”
Basically, the algorithm starts by looking at workload, which is scored based on weighing different kinds of assignments, INDs, and manufacturing supplements. The algorithm also looks at experience. “It’s a management- based ranking of experience,” Kozlowski confirms. “We ask questions such as, ‘Does this person have experience in this area? Have they reviewed a lot of antibodies? Have they reviewed a lot of enzymes?’” Management created a simple scoring system to rate that experience as one factor in the algorithm. According to Kozlowski, the math behind the algorithm has evolved significantly. “We have somebody who spends a lot of time gathering comments from customers on people and assignments so the algorithm can be constantly tweaked and properly weighted.”
During the development of the algorithm, Kozlowski had an idea on how to develop a similar approach to help OBP determine the best way to move people around within the new structure of DBRRs 1–4. “One of our challenges was having people buy into the new and more flexible organization that was created, independent of product structure and indication,” he says. “People really worry about their personal situation, such as who they are going to be working for.” Kozlowski put a lot of thought into how to move people around. For example, he first had discussions with all the managers, team leaders, or anyone who ran lab programs. He asked them a variety of general questions to get some idea of their expertise, which helped when determining how to divide them into groups. “I also asked them if there was one person they really wouldn’t want to work for,” he says. The idea was, whenever possible, to try to keep teams together. “That made moving people around much easier, because unless there was an issue, teams would be moved, not individuals.” The expertise data collected by Kozlowski was similar to what was being used in the algorithm, and he developed a simple scoring system that helped avoid the particular mismatches of what people claimed they didn’t want. “I came up with a number of models and then presented this spreadsheet to the leadership teams so they could visualize how we could reorganize,” he explains. Although he admits the spreadsheet models demonstrated there was really no way to strike a perfect balance among the divisions, they proved very helpful in other ways. “It’s a little tricky not talking about who wouldn’t want to work for whom,” he confides. “But I managed to do that by creating a number of models that I thought were balanced. So, when a manager would suggest moving a person to a different division, they could see how their proposal would change the scores and impact the balance. I don’t think I ever had to actually say, “I can’t do that because so and so doesn’t want to work for so and so.”
Kozlowski says the reorganization took more than a year, during which time workloads changed, people left, and new staff were hired, prompting multiple revisions to their calculations. “Still, all of that work on how to move people around created a lot of buy-in when the reorg actually happened.” Finally, though he realized that during the initial reorganization that one division may have a bit more experience in one area of expertise, he hopes with training and time, all the divisions will eventually even out.
Don’t Let Team Identity Become An Unanticipated Crisis
Steven Kozlowski, M.D., director of the Office of Biotechnology Products (OBP), says he and his leadership team had a lot of discussion on preparing for the eventual transition during the recently executed reorganization. In fact, they even distributed copies of the book, Managing Transitions by William Bridges, to members of the transition team to help them prepare for possible resistance. “One unforeseen outcome of the announcement of the reorg was that some employees felt they were losing their professional identity. For example, in the old division of monoclonal antibodies, the people there really identified themselves as being the antibody experts. But in the reorg, there would be antibody experts in multiple groups. In these kinds of situations, Kozlowski explained to the staff that they were victims of their own success. “Because the roles you played in helping facilitate these products made them grow so much that the system of doing this all in one division really won’t work anymore is a testament to your success.” In addition, he informed them that although their organization is losing its identity in the reorganization, OBP would still be tracking their individual expertise. “If you’re an expert in antibodies, for example, you’re still an expert in antibodies even if you work in a division that doesn’t have antibodies as its specific designation,” he explained. “Thinking about people’s identity issues and managing those issues is just as important as determining whom they’re going to work for and where their office is going to be,” he concludes.
The Importance Of Transition Teams And Working Groups
During a lot of the early parts where we were figuring out how to move people around and setting up a work structure for the eventual reorganization, the transition team was pretty much the current management,” says Steven Kozlowski, M.D. But the director of the Office of Biotechnology Products (OBP) at the FDA notes that as OBP got closer to the reorg actually happening, they took the opportunity to create a more rounded transition team. The original team included himself, his deputy director, Jeff Baker, Ph.D., and two current division directors. “But then we did interviews and actually had a selection process to add other people to that group,” he states. The goal was to make the process as transparent as possible. Eventually the group grew to nine members. “We thought it very important to have a strong cadre of appropriate leadership during the time of transition,” Kozlowski acknowledges.
It was this group that decided who the new acting division directors would be, as well as who would be acting review chiefs — a position that manages the full-time reviewers within the division. The transition team also played a critical role in determining what working groups were needed.
“Before the reorg, we had something called team leader lab chief meetings,” Kozlowski says. “If we felt there was an issue that was treated somewhat differently, technically, in one group than another group — say viral clearance for early studies of these products — we would discuss it there, and we would try to figure out whether the difference was based on a true technical scientific reason or whether we should try to be more consistent in how we approach that across the organization.”
While these meetings remain important and still occur weekly, the transition team realized there had to be a better way to share knowledge and best practices across four divisions that could potentially be reviewing the same type of product. The result was the creation of dedicated working groups that included representatives from all four divisions. These groups, often consisting of between eight to 10 people, tend to meet weekly, but the frequency can vary. “We have a working group for biosimilars, so that we can make sure we’re treating biosimilars in a consistent manner across the office, because there potentially could be two biosimilars to the same reference product being reviewed in two different divisions,” he explains. This can also occur with novel biological products. For example, on Aug. 27, 2015, the FDA approved evolocumab as an antibody to the target PCSK9 gene. Another antibody to the same target, alirocumab, was approved in late July 2015. “These two products, both monoclonal antibodies against the same target, were reviewed in two different divisions by two different teams,” Kozlowski explains. And while he views this as a demonstration that the reorganization is working, he also feels this exemplifies how working groups help to ensure consistency across OBP’s four new divisions. “We have a working group for immunogenicity, because again, while immunogenicity may be a different risk based on the nature of the product, it’s very important to have consistent ways of looking at that across the office,” he states. “Even though the adverse events for immunogenicity are a clinical issue, we play a large role in reviewing the assays for immunogenicity and important interactions to help determine if there is a link to adverse events.”
Working groups play a critical role in OBP’s strategy of having any division review any product. In fact, Kozlowski says they will be constantly re-evaluating whether additional working groups are needed.
Serendipity Plays Role In Removing Silos
During the process of reorganizing the Office of Biotechnology Products, its director, Steven Kozlowski, M.D., was faced with another big move, one that was actually a physical relocation. “The vast majority of OBP was located at the NIH, including our laboratory program, which is a very large and complicated effort,” he explains. The plan was to relocate OBP from the NIH Bethesda, M.D., location to the FDA’s White Oak campus in Silver Spring, M.D.. Though only a 25-minute move, Kozlowski says if you think moving people around is hard, moving labs is harder. “This led to a whole bunch of interesting challenges,” he shares. “We’re reorganizing. We have an idea about what the new divisions will consist of, but because of the way the rules work, we can’t really share the final org charts until they get approved. How do we move people into new offices when we can’t necessarily share what division they’re going to be in?”
For Kozlowski, this challenge turned out to be a bit of serendipity and an opportunity that fit nicely with the goal of OBP being less siloed. Here’s why: The reorganization of OBP was more a mental exercise that began in early 2013. “We had our idea of how 150 people would move during the crosswalk [a term used by government to denote the implementation of the reorganization] in early 2014,” he shares. “But the actual reorg didn’t happen until Jan. 15, 2015.” However, the physical move required a much more rigid timetable in order to be successfully executed during the summer of 2014. “We were all so busy with the physical move that while we were waiting to get the reorganization approved, people had a lot of other things to worry about beyond what division they would be in,” he notes. “Besides, why do we really need to keep divisions physically together anyway? If all the divisions can review similar products, maybe having your next-door neighbor being from another division where you talk to them and ask questions is actually a good thing.” Although he admits it was challenging to have all of these things going on at once, it turned out to be a real benefit. “Working together on one campus in a way that we weren’t before reinforced the idea of being part of one connected office,” he states. Kozlowski believes that everybody mixed together will be much more important to building a cohesive culture at OBP than relying on lines drawn on a formal organizational chart.