Novel Approaches to Vaccine Development

Speakers:
Philip Dormitzer, Novartis Vaccines and Diagnostics
Kanta Subbarao, National Institute of Allergy and Infectious Diseases
John Treanor, University of Rochester School of Medicine and Dentistry
Edwin M. Kilbourne, New York Medical College

Highlights

• The potential for antigenic change and evolution complicates vaccine development against pandemic H1N1.
• The conventional use of eggs to develop and produce vaccine could be limiting options in the United States.
• Many scientists are working with newer methods, including reverse genetics, to produce virus vaccine strains, baculovirus-expressed hemagglutinin, virus-like particles, fusion proteins, nanoparticles, and other potentially effective vaccine candidates.
• Newer adjuvants have the potential to increase vaccine potency and yields.
• Novel production technologies, such as using cell-based virus production, could speed vaccine development.
• A combination of approaches may be required to thwart the pandemic

New strain, tried-and-true approach

Once it was clear that a swine flu pandemic was looming, researchers geared up to develop an effective vaccine. Just as the public health community had been preparing for such a situation, vaccine developers were in a good position to do so, since work on an avian influenza vaccine has been underway for several years. Some of the promising strategies from those endeavors might also be applicable to the development of a swine flu vaccine, the presenters said.

The laboratory of conference coordinator Doris Bucher was among those galvanized by the H1N1 outbreak to develop seed strains for a potential vaccine. In a pre-conference interview, Bucher explained that influenza viruses come in three varieties—A, B, and C. Swine flu is type A and, like all type A flu viruses, it has a segmented genome that allows it to reassort, or mix, its genes with another type A strain if two viruses infect the same cell.

Public health officials are banking on the utility of a vaccine aimed at the current pandemic strain.

This potential for reassortment complicates vaccine development because currently three type A flu strains are circulating: the novel swine H1N1, as well as the H1N1 and H3N2 viruses that cause seasonal flu. ("H" stands for hemagglutinin, a protein that helps the virus get into the cell by binding to polysaccharide chains on the cell surface, resulting in entry of the viral genome into the cell. "N" stands for neuraminidase, molecules that help the virus particle leave an infected cell—and go on to infect other cells—by clipping off the ends of the polysaccharide chains.)

"If two type A viruses infect the same cell, the genes can shuffle and create new flu viruses. That's the concern," Bucher said. "You may have H1N1 and H3N2, and end up with H1N2." Novel combinations could potentially result in strains that are resistant to antivirals and unaffected by existing vaccines.

But despite the possibility of future reassortment, public health officials are banking on the utility of a vaccine aimed at the current pandemic strain. Four of the strains produced by Bucher's lab—which normally produces seed strains for the seasonal flu vaccine—were submitted to the Centers for Disease Control and Prevention (CDC); one strain looked particularly promising (NYMC X-179A) and was sent to manufacturers to evaluate and create seed lots.

Bucher's lab produces seed strains for the production of the inactivated virus vaccine the tried-and-true way—using eggs—and taking advantage of the virus's natural propensity to reassort. They infect an egg with an old virus, which is well adapted to growth in eggs, and the targeted new virus; after the strains reassort, they take various steps to ensure that no surface antigens from the old strain remain. The result is a vaccine candidate aimed at specific strains of the new virus. The strains, Bucher said, are "highly adapted" to grow in eggs, the conventional strategy for mass producing vaccine in the United States.

Live-attenuated viruses, reverse genetics

Kanta Subbarao and colleagues at the National Institute of Allergy and Infectious Diseases (NIAID)'s Laboratory of Infectious Diseases are also working on a pandemic vaccine—but of a different type, and using different methods. Subbarao's lab has been involved in pandemic preparedness against a potential avian influenza viruses, in collaboration with scientists at MedImmune, who have developed FluMist, a live-attenuated trivalent seasonal influenza vaccine.

Now the lab is using plasmid-based reverse genetics techniques to develop a live-attenuated vaccine (made from a weakened strain of the virus) for H1N1, rather than relying on genetic reassortment. The latter method requires that researchers sift through more than 250 possible combinations of viruses to find one that displays the desired hemagglutinin and neuraminidase antigens. By contrast, reverse genetics allows researchers to combine hemagglutinin and neuraminidase genes from a circulating virus with genes from a vaccine strain to generate the selected strains.

The effort is part of a larger pandemic influenza vaccine program that has been underway for about four years, Subbarao explained. The program involves the generation and evaluation of a library of vaccines against several influenza A subtypes; clinical trials at Johns Hopkins University to evaluate safety, infectivity, and immunogenicity of candidate vaccines in healthy adults; and banking of sera from vaccinated volunteers to test against bird and animal viruses that emerge in humans.

Live-attenuated vaccines have several potential advantages over inactivated vaccines, Subbarao said. Studies suggest:

• A single dose of a live-attenuated vaccine may be sufficient to elicit an immune response, whereas with an inactivated virus vaccine, at least two doses may be needed to elicit titers high enough to generate protection.
• A live-attenuated vaccine induces serum and mucosal antibody responses as well as T-cell responses, and could induce broader cross-protection than inactivated virus vaccines in naïve populations such as children.
• One embryonated egg could potentially yield 5000 doses of vaccine, compared with about a single dose of inactivated virus vaccine.

A vat of vaccine, with adjuvant

"It seems incredible that in the United States, we're still using eggs to make vaccines. No chicken, no eggs, no vaccines," marveled Philip Dormitzer of Novartis Vaccines and Diagnostics. Yes, the use of eggs—whether for inactivated or live vaccines—has been in widespread use for decades, has been proven safe, and doesn't require a steep learning curve. However, the low-tech approach requires a lot of lead time and advance planning.

Clinical trials are being set up to test how well adjuvants reduce dosage requirements.

By contrast, a Novartis manufacturing site in Europe is growing "a vat of vaccine" at a time. Dormitzer explained, "Clinical trials have shown that this cell-culture approach works well, with comparable efficacy to the egg-based vaccine" for seasonal influenza. Novartis is currently building a cell-based vaccine production facility in the United States, funded in part by the government. This means that the technology, which was licensed for use in Europe in 2007, will be ready to go when the factory is finished and agreements are in place.

Regardless of whether a potential H1N1 vaccine is grown in eggs or using other approaches, Dormitzer considers it likely that adding an adjuvant (an agent that enhances the activity of a vaccine but is not the antigen that the immune system is being trained to recognize) will reduce high-antigen dosage requirements and improve immunogenicity. Clinical trials are now being set up to test this concept for the H1N1sw vaccine. Adjuvanted vaccines with good safety records, he said, have the potential to:

• speed up vaccine availability, especially when used in conjunction with cell-culture technology
• rapidly boost a long-lived memory response to priming
• provide cross-clade immunity

However, despite the promise, "hundreds to thousands" of adjuvants have been described in the literature, but very few are in clinical use, Dormitzer stated. This has been due to difficulties in balancing efficacy and tolerability, manufacturing hurdles, problems scaling up, lack of reproducibility, and expense. The net result is that in the United States, the only approved adjuvant is alum (soluble aluminum salts), which has been in use since the 1920s.

By contrast, in Europe, the first oil-in-water emulsion adjuvant, called MF59, has been used in the Novartis's seasonal influenza vaccine Fluad since 1997, and more than 46 million doses have been distributed worldwide, Dormitzer said. Although the adjuvanted vaccine does cause more injection-site soreness than unadjuvanted vaccines, no increases in other adverse events—Guillain Barré syndrome or other autoimmune responses, cardiovascular events, hospitalizations, etc.—have been documented. Novartis has submitted data on the adjuvant to the U.S. Food and Drug Administration for evaluation, with a view toward developing a similar vaccine against H1N1.

The use of adjuvant is especially important in the context of a pandemic, Dormitzer observed. Titers are substantially higher with adjuvant, and smaller doses of antigen could be given, enabling manufacturers to produce many more doses from the same batch, compared with non-adjuvanted vaccines. Studies are underway to see whether priming with an adjuvanted vaccine and boosting later could generate more robust responses against the primary strain, as well as a broader response against other clades.

Dormitzer also proposed some novel solutions for overcoming another rate-limiting step in vaccine production and distribution; that is, the requirement that sheep be immunized and their sera used to test a vaccine's hemagglutinin antigen before a vaccine is released for distribution. This serum assay—single radial immunodiffusion—is the only accepted hemagglutinin antigen assay used to certify vaccines, but its use "can delay vaccine release by weeks," Dormitzer said.

"An emergency [such as the current pandemic] is an opportunity to make improvements," Dormitzer stressed. Several approaches could move things along, he suggested. Antiserum could be developed more quickly by using research-grade seed to immunize sheep, if an adjuvant were used to speed up antiserum production, if pig convalescent serum were used for swine-origin influenza, or if a non-serum-based assay were used to ensure standardization.

Trading eggs for insects

John Treanor of the University of Rochester School of Medicine and Dentistry also wants to change the fact that "roosters have to do their part in order for Americans to respond to a pandemic." His laboratory is working on an alternative approach that involves inserting an influenza hemagglutinin gene into a recombinant baculovirus and using insect cells to produce vaccine. The approach is "similar to what's done in cell culture," he said, "and tests of the resulting vaccine in healthy adults show about 80% efficacy in laboratory- or culture-documented influenza."

Using insect cells as a production technology has several potential advantages, according to Treanor. These include the fact that the approach is already in use for a human vaccine (HPV); protein production has been shown to be efficient and could yield increased supply; it's easy to scale up—and doesn't require specialized facilities; and there's no need to handle live virus, which eliminates some safety concerns.

Other approaches to vaccine development

A number of other non-traditional approaches to vaccine development and production are being investigated, Treanor observed. One such approach involves fusing the globular head domain of the protective hemagglutinin antigen with the Toll-like receptor 5 ligand, flagellin, and expressing the resulting fusion protein in E. coli. "The dogma was that it wouldn't work because bacteria don't process proteins the way eukaryocyte cells do. That's why my group started looking at insects [as a production mechanism]," Treanor said. But studies have shown that, contrary to expectations, the approach does generate a functional antibody response against an egg-grown virion.

Treanor's group has also developed a nanoparticle that could be used for immunization. The nanoparticle is a fusion protein made up of hemagglutinin and glycoprotein D, which self-assembles into a particle that displays hemagglutinin on the outside. Animal studies with H5N1 showed the nanoparticle is "highly immunogenic," he said.

A number of adjuvants are in varying stages of development, with some demonstrating "dramatically enhanced" immune responses and induction of antibodies across clades, Treanor said. These include a transdermal adjuvant patch that delivers a (nontoxic) heat labile toxin with vaccine, and CPG, an immunostimulatory protein (oligodeoxynucleotide) that seems to boost B-cell responses.

Various vectors for delivering antigens are also being tested, Treanor observed. These include adenoviruses (used extensively as vaccines in the military); poxviruses (for example, MVA [Modified Vaccinia Ankara], used for smallpox vaccination); and highly attenuated Newcastle Disease virus for mucosal administration.

DNA vaccines are also promising, in that they could be produced rapidly and relatively easily, and might generate cell-mediated immunity. However, difficulties in inducing high titers of antibodies in humans, the need for potent antigens or more complicated delivery approaches, and lack of demonstrated protective efficacy in humans mean these vaccines won't be ready for use any time soon, Treanor explained.

And, although cross-protective vaccines would be desirable, tests to date suggest the protective effect of any such vaccine would only reduce the severity of illness, not prevent it. Therefore, demonstrating protection in clinical trials will be "challenging," Treanor said. Moreover, it's not clear what role cross-protective vaccines might play in the context of seasonal influenza.

Balancing novel influenza vaccines for novel H1 viruses.

Adopting any new approach or combination thereof would involve, as Edwin Kilbourne showed, weighing of factors on both sides of a scale, according to Treanor. Factors such as the urgency of producing a vaccine, inadequate supply, and rapid antigenic evolution—which might argue for a novel approach—would have to be weighed against factors such as previous experience (with eggs, for example), existing capacity, predictability of response, and lead time.

For now, Treanor said, "I'm a believer in live attenuated vaccines and don't think the potential for reassortment is a major concern." But many people don't agree, he acknowledged, and several questions remain: How much shedding would occur; what doses are appropriate; and whether such a vaccine would be as effective in adults as it has been shown to be in children.

Learning from the past, and looking forward

In his keynote address, Edwin Kilbourne warned that, regardless of which strategies or combination thereof are implemented, it would behoove scientists and public health officials to remember lessons learned from the 1976 outbreak. At that time, Kilbourne emphasized, despite the often negative view of the National Immunization Program that for the first time was undertaken in 1976 to limit an emerging pandemic threat, many lessons in pandemic preparation were learned. Preliminary trials of vaccine dosage protocols were successfully carried out before the program began, and public health agencies at the national and state level were organized. Approximately 43 million people were immunized. Infectious disease experts also learned that the introduction of a virus of new subtype into a susceptible, crowded military population could disappear after successive person–to–person transmission and never spread into the general population. In hindsight, the rush to vaccinate everyone was probably unwise, he suggested, although prior experience had shown a decline and virtual disappearance of influenza virus infections in the summer months in earlier epidemics.

The termination of the program was triggered not only by the failure of the virus to spread beyond Fort Dix, but also by what appeared to be an undue prevalence of Guillain Barré Syndrome (GBS). This complication has been associated with many minor illnesses as well as many other vaccines, but, Kilbourne advised, it must be recognized as a possible complication of any future mass immunization program. Intense study of the X-53a vaccine used in the program demonstated no unusual neuropathic effects in experimental animals. In future programs the public must be adequately informed of possible risks from vaccination and any untoward effects should be compensated for. As with the post-Fort Dix program, preliminary studies of vaccine dosage should be conducted.

As of this posting, the H1N1 pandemic continues with no end in sight. Because the future remains uncertain, all potentially effective strategies to protect against and/or treat future cases are being pursued. As Treanor suggested, it is likely that a combination of approaches —e.g., inactivated vaccine as well as live vaccines—might be required to protect all vulnerable populations.