Breaking the Bottleneck
AIDS vaccine researchers and developers address the short supply of manufacturing and process development capacity
By Sheri Fink, MD, PhD*
There is a growing consensus within the AIDS vaccine research community that its various members need to work together to confront one of the major challenges to AIDS vaccine development—the acute shortage of vaccine manufacturing and process development capacity. For a candidate vaccine to move from the laboratory into clinical trials, and ultimately into commercial production, a process must be developed for its safe, reliable, well-characterized, cost-effective and large-scale manufacture.
After proposing a global AIDS Vaccine Enterprise (Science 300: 2036; 2003), leading vaccine scientists met last summer to begin establishing a strategic plan to accelerate AIDS vaccine development and six working groups were established to examine “needs, gaps, opportunities, and potential initiatives.” Two of these Vaccine Enterprise groups met jointly this February in Bethesda, MD, bringing together twenty world experts on vaccine manufacturing and process development from both the private and public sectors. The two working groups were the manufacturing group chaired by Jerry Sadoff (CEO, Aeras Global TB Vaccine Foundation) and R. Gordon Douglas, Jr. (freelance consultant on vaccines, infectious diseases and global health), and the product development group chaired by Gary Nabel (Director, Vaccine Research Center, U.S. National Institutes of Health) and Emilio Emini, (former Senior Vice President of Vaccine Research, Merck and Co., now with the International AIDS Vaccine Initiative). Their remit is to address the scarcity of AIDS vaccine manufacturing and process development capacity outside of large pharmaceutical corporations, shortfalls that are being blamed for causing delays in clinical trials. Although it will be some time before an effective AIDS vaccine is found, the experts agreed that planning for that day must begin now.
For vaccines, it is often said that the process is the product. Unlike most pharmaceuticals, vaccines are biological products created by manipulating complex systems such as mammalian cell cultures, embryonated chicken eggs or bacterial cultures. The vaccine manufacturing process is therefore subject to the fragility of these biological processes.
A failure to invest in process development may result in a vaccine that is not optimized for stability or yield, which could well affect costeffectiveness, all critical factors for a product whose greatest demand will come from the developing nations. This effort requires time, human capital, dedicated manufacturing facilities and, of course, funding. All are in short supply. The difficulty is that the vaccine field, which is already stretching finite resources for basic research and clinical trials, must allocate funds for the production of vaccines that do not yet exist.
There are two related problems. One is a shortage of manufacturing capacity sufficient to provide the quantity of vaccine needed for Phase III clinical trials. The second is a shortage in large-scale manufacturing capacity and space and expertise for process development that will enable scale-up to maximum production once an efficacious vaccine is licensed. As a rough approximation, the process needs to leap from the 200 ml laboratory flask used to produce the initial vaccine to the several hundred or thousand liter bioreactors needed for clinical trials, and then to the 10,000 to 50,000 liter reactors required to make hundreds of millions of doses of vaccine during full-scale manufacturing of an approved product.
The field has reached a critical moment because of timing. The five or more years required to design, build and license a largescale manufacturing facility approximates the number of years required to bring a vaccine candidate through a Phase III clinical trial. “That means you need to make your major investment at the time you’re starting your Phase III trial,” says Sadoff. “That’s a problem because you don’t know if [your product] is going to work.” The demand (as distinct from the need) for the vaccine is another unknown that will depend on factors such as efficacy, duration of protection, immunization schedule, safety, cost, and stability, the latter two of which are, in turn, influenced by the degree of early investment in process development. Adding to the risk is the question of who will pay for the vaccine in the many low-income countries where it will be needed most.
On the other hand, with more than two dozen vaccine candidates currently in Phase I or II trials, a failure to make immediate, significant investments in process development and manufacturing capacity will result in delays in bringing a vaccine to market should any of these candidates be found efficacious. When it comes to HIV, those delays will be measured in lives. “Anything you can do to minimize time rather than resources and risk is key. You’re talking about millions of lives” says Sadoff.
The state of the field
Research scientists developing AIDS vaccine candidates outside of major pharmaceutical companies frequently rely on contract manufacturing companies to provide clinical-grade materials for early clinical trials. Because the quantity of vaccine required for a Phase I safety trial is small, little if any manufacturing scale-up is necessary, and the process used by the contract manufacturer is similar to the process used to make the candidate in the original research laboratory.
Even so, scheduling outside contractors for a Phase I trial is often difficult. “One great problem is these contract manufacturers are constantly booked; you have to pay a reservation charge and book your slot [well in advance],” says Eddy Sayeed, Vaccine Production Manager for IAVI. “If one is not able to achieve the timelines, you lose your slot and have to pay a certain amount for not being able to fulfill the requirement.”
As much of a problem as these logistical issues pose in early phase clinical trials, vaccine developers face an even tougher question: What to do once they reach Phase III trials and ultimately the need for fullscale production? Outside of the large pharmaceutical companies, there are no pilot, large-scale manufacturing plants for Phase III production capacity of viral vectors, which, along with DNA and protein-based vaccines, are the three major types of current vaccine candidates (see Figure 1). Such plants and the expert workers that go with them are needed not only to manufacture vaccine for trials, but, perhaps even more importantly, to develop an optimized process for large-scale production.
Figure 1. The three types of manufacturing technologies used to produce different AIDS vaccine platforms.
“The process development issue is critical,” says Emini. “Process development requires intellectual and physical investment which is housed at the moment almost exclusively within the large biopharmaceutical companies.”
Nabel agrees. “If you don’t have a manufacturable process and if you don’t have vaccine candidates that can really be manufactured to the specifications and the scale that you need, then you don’t have a vaccine.”
The shortage of process development and manufacturing capacity has led to a bottleneck and, some argue, a slowing of the entire vaccine development pipeline. “People are discouraged from making products if they’re worried there isn’t a capacity to do trials,” said Sadoff.
One AIDS vaccine candidate has progressed through all clinical trial phases, VaxGen’s AIDSVAX, a recombinant gp120 protein vaccine. While the mammalian cell culture manufacturing process used to produce recombinant proteins was fairly well characterized, and the process for gp120 had been developed by Genentech and licensed exclusively to the company, VaxGen still had to respond to the potential need for commercial manufacturing capacity. “We were faced with the fact that there was not anywhere, even projected in the future, capacity to produce recombinant proteins even on a moderate scale,” says Don Francis, founder of VaxGen.
In February, 2002, during Phase III clinical trials of AIDSVAX, VaxGen and a group of South Korean partners invested US$113 million in building a largescale biopharmaceuticals manufacturing facility in Incheon, South Korea, with an initial 50,000 liter bioreactor capacity, expandable to 150,000-200,000 liters. The joint venture also funded the construction of a smaller facility in San Francisco, US, to be used to validate the AIDSVAX manufacturing process. The two facilities were expected to produce 200 million doses of AIDSVAX per year with the completion and licensure of the South Korean facility in 2005 or 2006. Experts outside of VaxGen have complimented the company’s approach to developing large-scale manufacturing capacity for its product, which ultimately failed to protect against HIV infection. VaxGen is now recouping some of its investment by putting its US manufacturing capacity to work to produce other recombinant protein products, including anthrax vaccine. However, the unfinished South Korean facility has yet to find a new use, demonstrating that early investment in manufacturing capacity does indeed carry risk. “It’s not that the demand is not there,” says Francis, who believes the facility will eventually be put to good use once others are convinced that high quality vaccines can be produced in places such as Asia. “It’s the first recombinant [protein] manufacturing facility outside the US or Europe.”
The search for capacity
AIDS vaccine researchers are working to define the requirements of the field and the capacity that exists worldwide for developing and manufacturing various types of AIDS vaccines, including DNA, viral vector and protein-based vaccines. Early expectations were that major pharmaceutical companies in the industrialized world, which have the expertise and the facilities to undertake process development and largescale vaccine manufacturing, would take the lead in producing AIDS vaccines. However, it has become clear that these companies do not have a financial interest in tackling the project by themselves. “Companies are staying away from making an HIV vaccine because nobody knows how to do it,” says Douglas, “the science isn’t there.” Even if an efficacious candidate emerges, AIDS vaccine manufacturing may ultimately be complicated by concerns over product liability, intellectual property, and pressure to make early technology transfers into the developing world.
Still, research and development experts from the large pharmaceutical companies of the industrialized nations are active participants in the current discussion about AIDS vaccine production. Companies may end up contributing in very valuable ways, such as offering training in process development and manufacturing, or making their validated cell lines and egg banks available to AIDS vaccine developers.
Emini envisions that large biopharmaceutical companies might have an interest in manufacturing an AIDS vaccine once it has been proven effective. This would still leave the need for early process development unmet. “Process development has to be done up front,” he says, “that’s a key gap.”
While large pharmaceutical companies may be willing to manufacture an AIDS vaccine for developed country markets, where profitability or at least return on investment is possible, producing for developing countries is a different question altogether. Yvette Madrid, a freelance vaccine policy consultant, has studied the issue: “For a commercial firm, there’s almost no incentive to build a huge amount of capacity for the developing countries, because that increases their risk; there’s the possibility that the supply won’t be used, or the demand won’t materialize.”
These problems could be mitigated if a donor stepped forward with a large, early, and firm commitment to buy an AIDS vaccine at full cost and provide it affordably in the developing world. While a lack of funding to cover many other needs in AIDS vaccine development may make this possibility unlikely, technology could realistically be transferred to lower-cost manufacturers. “One of the solutions,” says Douglas, “is to have [the vaccine] manufactured in some of the middle-income countries like Brazil or India, where it can be manufactured at lower cost because labor costs are lower.” The manufacturing facility for AIDSVAX in South Korea is a case in point.
Building a multi-product clinical trials manufacturing facility
A multi-product clinical materials plant capable of producing 10 or 20 products simultaneously would open the AIDS vaccine development pipeline. If excess clinical materials manufacturing capacity existed, the attitude to testing new vaccine candidates in humans could change, encouraging both a broadening of the search for novel candidates and the testing of finer variations of promising vaccine platforms that differed only in their immunogens.
Various models are currently under consideration. These range from a facility dedicated solely to the production of clinical materials to one that would include full process development capability and the capacity, with additional investment, to be scaled up to commercial 10,000 liter vaccine production. Cost estimates for the different facilities vary, from $50 million to $200 million.
One useful model is the $65 million pilot plant being built in Frederick, Maryland by the Vaccine Research Center of the National Institutes of Health to produce AIDS and biodefense-related vaccines. The plant, a year away from completion, will contain four manufacturing suites of various capacities (two 100 liter, a 400 liter and a 2,000 liter) that will be capable of meeting current Phase III needs. “It’s a nice design,” says Nabel, “it’s flexible and it allows for different scale production.” While this plant will be occupied for the foreseeable future with NIH products, the plans for the facility are being shared openly with the Vaccine Enterprise.
The future of full-scale vaccine manufacturing is also being considered. “The object of this exercise is to identify the end point, a manufacturing facility,” says Don Gerson, Managing Director of Manufacturing at IAVI. “How big is it going to be, what does it have to make, if we’re really going to supply the entire developing world with an AIDS vaccine?” For construction to commence before clinical trials have shown which type of vaccine will be needed, a flexible facility could be built that would allow, with slight modification, for the use of any of the three major types of production processes—egg-based, microbial-based or cell culture-based. “It’s very easy to design for all three possibilities at once,” says Gerson. “If you design just for one, and you go back and say ‘it should have been eggs instead of cell culture,’ you have to start all over again.” Alternatively, Phase IIb clinical trials could be performed prior to construction, providing proof of concept; the merits of Phase IIb trials are currently being debated (see “IIb or not IIb”).
While the need for a high-efficacy AIDS vaccine in the industrialized world would probably fall under 100 million full immunization courses, the goal of covering estimated world needs for such a vaccine could possibly take peak requirements close to 700 million immunization courses (Vaccine 21: 2032; 2003). This could require the work of four to five regional manufacturing facilities, each possibly costing as much as $300 million to build.
Currently, vaccine manufacturing capacity in general is finite and already causing shortages of licensed vaccines for other diseases. “The only way to make more [vaccine] is to build another plant, and that’s going to take five years,” says Douglas. “If you don’t have the demand situation worked out—not the need, but actual demand—you’re not going to know what size to build for. If you build too big, you raise the cost prohibitively on the vaccine; if you build too little, it’s going to be a crisis. We have to get into these issues now.”
“The challenge for the world is to think about how to get to large-scale manufacturing as soon as possible,” says IAVI President Seth Berkley. Across the AIDS vaccine field, the importance of addressing shortages in process development and manufacturing capacity has now been acknowledged. “Researchers in the field recognize the need for coordination, and this is the time to do it,” says Emini. “These are problems that go beyond what any single individual can do, what any single organization can do.”
*Sheri Fink, M.D., Ph.D., is a freelance writer whose work has appeared in such publications as the New York Times and Discover Magazine, and the author of "War Hospital: A True Story of Surgery and Survival.”