Abstract
The Informa Life Sciences vaccines conference is an annual meeting of a relatively small number of academics and industrialists. It is split into three concurrent sessions covering vaccine discovery, quality and manufacturing. Although there were many presentations of merit, only a few will be discussed here, including the plenary speeches on adjuvants and influenza.
Trends in developing new adjuvants
There is an increasing trend to move from inactivated or killed whole-pathogen vaccines to subunit vaccines. There are two motivators for this. The first is the determination of the specific antigen required to generate an immune response and the second is the desire to reduce the reactogenicity of the vaccine so that it may be considered ‘safer.’ However, in doing so, Gary van Nest (MedImmune, MD, USA) explains that we are reducing the potency of the vaccine. This is because many of the old, whole-pathogen vaccines had a self-adjuvanting capacity based on the presence of certain Toll-like agonists. For example, bacterial flagellin, lipopolysaccharide and DNA are known Toll-like receptor (TLR)5, TLR4 and TLR9 agonists, respectively, which stimulate the innate immune system. Subunit vaccines are generally formed with a few purified proteins that are known to be antigenic, thus excluding these TLR agonists. Hence, the protein dose needs to be increased in order to stimulate the immune memory, but even this may not be long-lasting. However, there are other drivers for new adjuvants: modulation of the immune response (Th1 to Th2), the need for enhanced dose-sparing properties for difficult-to-manufacture vaccines and enhancing the cell-mediated immune response.
Thus, a modern adjuvant for subunit vaccines must be able to facilitate the delivery of the antigen to the immune system, as with alum, and be immunopotentiating to increase potency. This will require a systematic approach to adjuvant research, identifying both the immune stimulators and characterizing the receptors activated in the innate immune system. Immune stimulators range from lipids, CpG and saponins to RNA and DNA. The receptors that they stimulate are not necessarily restricted to TLRs, but can also include the nucleotide-binding oligomerization domain and retinoic acid-inducible gene-1.
Since the 1940s, aluminum salts have continued to be the preferred adjuvant, as early work with oil-in-water emulsions, such as Freund’s adjuvant, led to adverse reactions in humans. However, research continued in this field and new emulsion adjuvants have been licensed. Examples of this are MF59® (Novartis, Basel, Switzerland), AS03 (GlaxoSmithKline [GSK], Belgium) and AF03 (Sanofi Pasteur, Lyon cedex, France). Furthermore, new particulate antigen carriers have also been developed, such as liposomes, micorparticles, virus-like-particles (VLPs) and other immunostimulating complexes. These provide more efficient delivery to antigen-presenting cells and facilitate antigen display, codelivering both antigen and adjuvant, and are thus an essential component of any modern adjuvant system.
Currently, there are four new adjuvants licensed by the EMEA. MF59 and AS03 are both oil-in-water adjuvants used in influenza vaccines within the EU only. AS04 (GSK) uses monophospolipid and is licensed for use in the EU (hepatitis B vaccine) and the USA (human papillomavirus vaccine). Virosomes, which are phospholipid vesicles incorporating influenza antigens, are licensed in several countries (the EU, Asia and South America) but not in the USA. So far, the US FDA has only approved AS04 for use with Cervarix™ (GSK).
Many more adjuvants are in development and in clinical trials, including CpG (Dynavax Technologies, CA, USA and Novartis, NJ, USA), GLA (Immune Design Group, WA, USA) and flagellin (VaxInnate Corporation, NJ, USA). Other stimulators include QS21, a saponin plant extract developed by Wyeth (Berkshire, UK) for Alzheimer’s disease and LT, a heat-labile enterotoxin from Escherichia coli developed by Intercell AG (Vienna, Austria).
The challenge for all these new adjuvants is the regulatory hurdle, as the risk–benefit ratio must be finely balanced. The EMEA states that adjuvants must have an optimal response and minimum side effects where the benefits override any harm. The FDA, however, sets the bar higher with a more cautious approach, demanding that no adjuvant can be introduced into a vaccine unless there is satisfactory evidence that is will not adversely affect the safety or potency of the vaccine.
Hence, the requirement for any new adjuvant will be to clearly demonstrate its benefit. Regulators will want to know the adjuvant’s method of action, whether it enhances the immunogenicity of the vaccine, reduces time to protection and improves efficacy, or if it has dose-sparing properties. This places a higher burden on adjuvant developers, which cements the need for mechanism-based research.
Influenza vaccines
H1N1 dominated the first day of the conference and for obvious reasons. With the second wave of the pandemic ongoing in several parts of Europe, manufacturers took time to champion their novel methods of vaccine production. There are several manufacturers of influenza vaccines, most notably GSK, Sanofi Pasteur and Novartis, and they all rely on fertilized eggs as the primary method of manufacture for the production of the subunit vaccine. However, owing to the limitations of the egg production method, most notably the inability to rapidly scale-up and maintain process monitoring, alternatives were sought. All three manufacturers are developing a cell culture capacity in which the viral propagation stages take place, but alternative technologies are also being investigated.
The Novartis method uses detergent extraction to create a purified subunit vaccine of 15 µg of hemagglutinin. An alternative is the use of VLPs reported by John J Trizzo (Novavax, MD, USA). VLPs consist of the virus capsid created in the absence of any genetic material. The advantage of this is a biosafety level 1, replication-incompetent virus vaccine using a baculovirus expression system. The production method consists of growing insect cells to a high cell density and then infecting them with an engineered baculovirus that produces the VLP of choice. Another component of this technology platform is the use of disposable production methods. Most production methods use large stainless steel vessels and thus lead to centralized production facilities. Instead, Trizzo proposed the use of a smaller disposable manufacturing facility that is both cost effective and productive. In comparison with existing technologies, Trizzo states that an egg-based production facility that produces 100 million doses a year on average requires a footprint of 140,000 feet squared (ft2) and costs US$150 million to construct. Compare this with a disposable facility that produces 75 million doses annually, has a footprint of 55,000 ft2 and costs $40 million to construct. Their models suggest that when applied to a pandemic scenario, a 10,000 ft2 facility will be capable of producing 25–30 million doses and will take 120 days to build, costing US$5 million. The appeal of this platform technology is to countries that do not have a background in vaccine production as the disposable platform has low infrastructure requirements. This means that these countries would be able to offer a vaccine to its citizens within months of establishing a manufacturing plant.
Novavax has already received $6 billion in finance over 4 years and has an aggressive commercialization strategy based on seeking regional partners. Currently they are partnered with ROVI SA in Spain and CPL Biologicals for operations in India. The company is still seeking regional partners in the Middle East, China and South America. By exploiting regional partnerships, it is possible to accelerate the path to licensure from the regional regulatory authority and achieve rapid expansion into the global market. Their product portfolio focuses heavily on influenza, both seasonal and pandemic. The seasonal influenza vaccine had passed Phase IIa trials, which enrolled 240 volunteers, and Phase IIb trials are about to start. Their H1N1 candidate vaccine has completed an initial trial in Mexico involving 4000 volunteers.
Novavax is not the only baculovirus-based manufacturer of influenza vaccines. FluBlock® (Protein Sciences Corporation, CT, USA) is a recombinant hemagglutinin and neuraminidase vaccine produced in a baculovirus expression vector system (BEVS). Daniel Adams (Protein Sciences Corporation, CT, USA) reports that their BEVS platform technology can be manipulated to produce any recombinant protein. The key to its productivity are the specially developed insect cell lines that are not based on SF-9 cells. These engineered cells are larger with thicker cell walls and are considered to be more productive. The BEVS system is nonadaptive, can be used in a biosafety level 1 facility and requires 2 months from strain identification to release the first batch of vaccine. FluBlock requires more protein than other seasonal vaccines (45 µg per strain instead of the usual 15 µg) but it does not use an adjuvant. More than 3000 people have been tested in large field studies in the 18–64 years age group and in the over 64 years age group. FluBlock has completed Phase III trials and met 21 of the 24 criteria set by the FDA. However, the FDA has recently requested additional safety checks after the review panel of 11 individuals split six–five in favor of further studies after one of the volunteers was hospitalized during field trials.
Although this is a setback for the Protein Science Corporation, their platform technology BEVS has been applied to other vaccines for prostrate cancer and diabetes. In funding, they received $8 million in 2008 and $9.5 million in 2009. They have secured a $147 million contract from the Biomedical Advanced Research and Development Authority (US Government) for their pandemic influenza vaccine PanBlock™, which has just entered clinical trials.
VLPs as antigen scaffolds
Since the development of the first VLP vaccine, research has continued to investigate other applications. VLPs offer a unique advantage over subunit vaccines as they also contain no genetic material, and thus cannot cause disease, but maintain the virosome structure so that it can acts as a TLR agonist, hence incorporating self-adjuvanting properties.
The Scripps Institute (CA, USA) was one of the first organizations to attempt to exploit the self-adjuvanting properties of VLPs by using a flock house virus as a scaffold with which to present antigens to the immune system. Mike Whelan (iQur Ltd, London, UK) elaborated further with his research into tandem-core VLPs using the hepatitis B (HepB) core protein. The HepB core protein spontaneously forms a highly immunogenic, self-adjuvanting VLP. The core protein itself forms spikes that are large enough to carry an insert; but large hydrophobic inserts cause the VLP to disassociate. However, by creating a genetic construct that links two spikes together into a single protein, a stable dimer is created that can accommodate large inserts. The resulting VLP is stable and has two antigen insertion sites. iQur’s lead candidate is a combined hepatitis A/B vaccine. In antigen insertion site 1 is the 88-kDa HepA virus P1 protein and in the second is the HepB surface antigen, 4.8 kDa in size. The advantage of this vaccine is that it requires only one manufacturing process, thus halving the cost of goods. The current combined HepA/B vaccine manufactures each component separately and then combines the two antigens in a final formulation. Work has already begun to produce enough material for clinical trials in parallel with research into optimizing the production process.
However, their proof-of-concept model also generated much interest. A dual construct of HepB surface antigen and M2 was created. M2 is an ion channel protein in influenza, the ecto-domain of which contains a highly conserved sequence of 24 amino acids. It is considered to be a lead candidate in the ‘universal’ flu vaccine. While the ELISA data detected a clear response to both surface antigen and M2, it was found that the presence of alum greatly enhanced the M2 response in mice serum. In mouse challenge models, mice were initially inoculated with the dual insert VLP (surface antigen and M2) with alum and infected 28 days later with influenza (PR8 strain). The mice lost weight over the first 5 days postinfection, but by day 10 they had recovered almost all the weight lost. Furthermore, the vaccine provided protection to 75% of the sample population. Although this is very early, the initial data suggest that the tandem core platform could be applied to other vaccines, including a universal influenza vaccine using M2.
New opportunities in VLP processing
The ability to quickly determine critical process parameters for manufacture of a novel virus and VLP vaccine candidates was a recurring theme. Several companies championed new chromatography systems using either monoliths or membrane systems. Aleŝ Ŝtrancor (BIA Separations, Ljubljana, Slovenia) spoke about a new monolith system, which utilizes a single piece of resin/matrix with an open pore structure to overcome diffusion limitations seen in bead-based resins. An alternate technology developed by Natrix Separations (Ontario, Canada) uses a chemically treated membrane surface with an appropriate ligand that can preferentially bind target antigens and purify in a manner similar to chromatography. Both technologies claim greater throughput than conventional bead-based chromatography and can be used in a disposable platform.
Nigel Titchener-Hooker (Department of Biochemical Engineering, University College London, UK) outlined scale-down models of both chromatography and primary recovery for VLP processing. Through the integration of statistical tools such as factorial design with these scale-down models, it was possible to screen multiple factors in a very short period of time, allowing the identification of the main factors and interaction effects that may occur between primary recovery and chromatography. The chromatography scale-down model consists of a pipette tip filled with resin using a working volume of 40–80 µl. The tip is attached to a robotic liquid handling system (Tecan) and the liquid is drawn up and expelled through the resin in both directions. After loading, the wash and elution steps follow and product recovery is monitored. Using such a device, it is possible to screen multiple resins, identifying the most suitable; but by maintaining a constant linear velocity, it is possible to accurately scale-up as well.
Hepatitis B surface antigen was used in this model as an example of a complex lipid protein VLP that is often difficult to process. The studies identified important interaction effects between the use of detergent for extraction of the VLP from cell debris post-homogenization and product loss. Furthermore, the scale-down platform permitted experiments into process modifications. In this case, the addition of a centrifugation step after cell disruption that carried the product in the solids fraction improved product yields at chromatography by 36%. It also decreased protein and lipid contaminants by 94 and 78%, respectively. Although this platform was applied to the HepB surface antigen, it could be applied to other virus and VLP vaccines and quickly provides process data.
Identification of potential anticancer antigens is key to creating a vaccine
Mai-Britt Zocca (DanDrit Biotech A/G, Norway) spoke on the development of their lead polyvalent cancer vaccine, MelCanVac®. There is a large number of shared tumor antigens that are expressed by multiple types of tumors. The key is to identify and extract antigens that will not lead to autoimmunity. ‘Cancer-testis’ peptides are a class of tumor antigens that are expressed in many tumors but not in normal tissue. The only exemption is reproductive cells, but because reproductive cells are not normally exposed to the immune system, cancer-testis peptides are considered tumor specific. Thus an allogenic melanoma cell line that expresses a wide range of cancer-testis peptides was utilized to produce antigens against cancer. These cells were homogenized and the resulting cell lysate was purified and loaded into dendritic cells for use as a cancer vaccine. While the melanoma cell line is engineered in house, the dendritic cell was prepared from the patient. A total of 200 ml of blood was extracted from the individual, and monocytes were isolated and converted into dedritic cells. The dendritic cells are then exposed to the tumor lysate containing the cancer-testis peptides and allowed to mature. This mature dendritic cell is then reinjected back into the patient to create a T-cell-mediated response against cancerous cells. MelCanVac has been tested in several clinical trials against colorectal cancer (CRC) and non-small-cell lung cancer (NSCLC).
In Denmark, metastasing CRC patients were injected intradermally with the product and tolerability and toxicity were monitored in an open Phase I/II clinical study. The product was found to be well tolerated in the 20 patients enrolled. Furthermore, in 20% of the group, the disease stabilized. A second Phase II study was conducted in Singapore and volunteers with advanced CRC were recruited. Out of the 20 patients, six reported that the disease had stabilized and one reported a reduction in tumor mass.
With NSCLC, 50 volunteers with inoperable lung cancer for whom chemotherapy options had been exhausted were recruited in Copenhagen, Denmark. Approximately 40% have responded well to the treatment, although the correlate of protection remains inconclusive.
While these data are very promising, it is the creation of a commercial GMP process that now poses a new challenge. The company has had a positive pre-Investigational New Drug meeting with the FDA, but creating and maintaining a robust, reproducible and cost-effective manufacturing process is not a trivial process.
Other talks of note
Elisabeth Linder (Diamyd Medical AB, Sweden) reported on a new treatment which seeks to preserve β cells and create a vaccine against autoimmune diabetes. Ron Ellis (NaxVax Ltd, Israel) spoke about their lead monoclonal antibody candidate blocking b-site 1, which inhibits plaque formation. This is a possible treatment for Alzheimer’s disease.
Conclusion
Although influenza appeared to dominate many of the talks, it was encouraging to see research into other diseases, in particular cancer. It is also interesting to observe that many of the influenza vaccine manufacturers were also seeking to diversify their portfolio and investigate other virus vaccines, such as respiratory syncytial virus. While investment in vaccines continues to grow, companies appear to be de-risking their businesses by creating manufacturing technology platforms that can accommodate any new vaccine antigen. However, with each new vaccine the methods and mechanisms to create a cost-effective manufacturing process may prove to be a limitation. Thus the importance of scale-down tools to reduce the process development time and rapidly identify critical manufacturing process parameters is highly valuable. While there do not yet appear to be any generic trends, one thing is clear – each vaccine is different, each with it own unique hardships.
Financial & competing interests disclosure
The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.