Vaccine Manufacturing Reborn

Adjuvants have a substantial impact on vaccine efficacy and in conserving manufacturing capacity.

Based on its burgeoning clinical-stage pipeline, GlaxoSmithKline (GSK; Rixensart, Belgium) expects its vaccine sales will at least double over the next 10 years. The company is banking on its Cervarix vaccine (against human papillomavirus, a cause of cervical cancer) to drive sales near-term. Several products currently in the pipeline are expected to reach the market over the next several years. In 2005, GSK committed itself to launching five new vaccines in five years, beginning with Cervarix and its Rotarix anti-rotavirus product. The company also is focusing on products for influenza, meningitis, oncology and allergy. All told, the company has 33 vaccines in preclinical through Phase III/filed, nine of them in late-stage testing. GSK expects at least half of its vaccine sales growth to accrue from vaccines currently in the development pipeline.

“Our biggest challenge in continuing our growth will be expanding our manufacturing capacity over the next four to five years, and obtaining approvals for the many vaccines within the major markets which are increasingly global,” comments John Picken, VP for industrial operations, North America at GSK Biologicals (Laval, Québec). It is common for vaccine manufacturers to seek approvals simultaneously in their countries of domicile and overseas.

GSK believes its future vaccine successes lie with novel, biological adjuvants. Pipeline products containing these new immunostimulants include Cervarix, which uses a biological adjuvant produced through fermentation. The adjuvant used in Cervarix is already incorporated into another vaccine marketed in Europe.

Although biologically “adjuvented” vaccines are approved as single products, each component commands its own manufacturing and purity controls, and the adjuvants must pass regulatory muster. Even though their manufacture is straightforward, biological adjuvants are unfamiliar to regulators as well as to their developers.

The impact of adjuvants on vaccine efficacy and in conserving manufacturing capacity is substantial. Early vaccines for pandemic flu lacking an adjuvant required two 90 microgram doses to induce an immune response. By adding adjuvant GSK vaccine reduced the effective dosage to 3.8 micrograms. “If the vaccine industry had to produce a flu vaccine that required 90 micrograms per dose we would never have enough capacity to supply the market’s needs,” Picken noted. GSK observed a similar effect with its malaria vaccine. Sixty-five percent of GSK’s vaccine pipeline depends on adjuvant systems.

Pfénex rising

At one time the word “adjuvant” was synonymous with alum and related aluminum-based compounds whose mechanisms of immune stimulation are still only poorly understood. Today, vaccine manufacturers increasingly look to biological adjuvants for their innate ability to jump-start the immune response to vaccines. Vaccine developers employ adjuvants strategically to increase vaccine efficacy, sometimes quite dramatically, and thereby lower the production cost per dose. In many cases vaccines simply do not work without an adjuvant.

Dowpharma has seen its adjuvants business triple since 2005, due in large part to the success of its Pfénex (pronounced “phoenix”) platform protein expression technology. The company expects high growth to continue based on existing projects.

Because it uses bacteria rather than animal cells to produce vaccines, Pfénex requires relatively short development and scaleup times. Kurt Hoeprich, commercial director for biopharma at Dowpharma, claims many protein-based adjuvants are difficult to express in traditional cell culture. And since they contain no animal pathoges such as viruses and prions, bacteria-derived adjuvants are ideal for injected vaccines. “Developers will use bacterial expression systems if at all possible because of speed, cost of goods advantages,” Hoeprich says.

Along with Pfénex, Dowpharma employs its “rapid strain identification” method to evaluate the manufacturability of a vaccine in a given strain of Pseudomonas fluorescens (the Pfénex “bug”) in eight weeks or less. “We can examine so many different ways of making a target simultaneously, that our odds of finding a high-yield system in eight weeks are extremely high.” Hoeprich credits the ability of rapid strain identification to assess up to thirty different production strains simultaneously as the “driving force” behind the success of Pfénex. All told, a manufacturing-worthy process takes about six months to develop, during which time Dow provides preclinical-grade material. The process is then licensed back to the customer. Dowpharma operates sixteen pilot-scale fermenters at its San Diego, California location but performs no GMP manufacturing.

During this past year, the first Pfénex-derived protein antigen entered U.S. clinical trials and should be in Phase II by the time you read this.

Dowpharma’s flagship customer continues to be vaccine developer Iomai (Gaithersburg, Md.), which went public in 2006. Iomai champions needle-free, patchbased vaccine delivery and has products in early clinical testing for traveler’s diarrhea, pandemic influenza, and seasonal flu. Iomai’s patch-based vaccines are stable at room temperature for up to six months, which means no cold-chaining or refrigeration.

Pfénex’s principal limitation is the inability of bacteria to glycosylate proteins or, more generally, to produce polysaccharides – an important class of antigens in their own right. So while Pseudomonas cannot produce glycosylated monoclonal antibodies, it excels at bacterial antigens that are almost exclusively unglycosylated. For polysaccharide antigens Dowpharma can use Pfénex to optimize a suitable protein carrier.

Dowpharma’s virus-like particle (VLP) antigen carriers, also generated through Pfénex, possess the size, shape and surface characteristics of a pathogenic virus but are non-infective. Dow engineers antigens directly into VLPs, during expression within Pfénex, so chemical conjugation is not necessary. The company has recently won a CRADA from the National Institutes of Health to develop an anthrax vaccine using VLPs. Primate studies have shown this approach to be promising. As with Pfénex, Dowpharma’s business plan for VLPs is based on licensing plus royalties.

HEPLISAV, a Phase III hepatitis-B vaccine from Dynavax (Berkeley, Calif.) uses a proprietary oligonucleotide immune stimulator, 1018 ISS. HEPLISAV vaccine induces seroconversion (antibody generation) in 100% of subjects, according to a Phase II study report released in December, 2006. Using a similar vaccination schedule, the GSK hep-B vaccine, which uses an alum adjuvant, induced seroconversion in only 70% of subjects. The difference, explains Stephen Tuck, who manages biologicals production at Dynavax, is most likely due to the 1018 ISS adjuvant.

Given that Merck and GSK already have marketed hepatitis B vaccines, the challenge to Dynavax is to add value to the vaccine, manufacture it cost-effectively and achieve these goals in a way that satisfied regulators in Asia, Europe, the U.S. and Canada – all major markets.

According to Tuck, EU’s Clinical Trials Directive 2001/20/EC, which aims to protect patient rights during clinical trials, has hampered foreign drug makers from testing their products within the EU. So, instead of seeking entry of HEPLISAV through a circuitous route involving an EU-certified “qualified person” who would vouch for Dynavax manufacturing, the company purchased Rhein Biotech (Düsseldorf, Germany). This plant, which was already licensed to release clinical trial materials for testing throughout the EU, also had experience manufacturing the specific hepatitis antigens Dynavax needed for the vaccine. If getting a foothold into European markets was the only goal, purchasing the facility would make no sense. But here Dynavax gets the plant plus an entrée into EU testing. “We got a package,” Tuck notes.

Specifically, Rhein Biotech had invented a proprietary technology for manufacturing the antigens in a yeast, Hansenula polymorpha, which produces the antigens in very high yield. This particular yeast is methylotrophic, which means it utilizes highly efficient C-1 carbon sources such as methanol. The Hansenula manufacturing system has been licensed throughout the world to Sanofi Pasteur, Green Cross (Korea) and others, and has been the basis for the manufacture of more than 900 million doses of hepatitis B antigen.

When it is approved, the hep-B vaccine containing 1018 ISS will only be the third oligonucleotide drug licensed in the United States, and the first to employ phosphorothioate base linkages. You may recall this technique was devised in the 1990s to confer stability to antisense molecules. 1018 ISS is manufactured through solid-phase synthesis – no fermentation or cell culture – by Avecia Biotech (Boston, Mass.), an AstraZeneca spinoff.

Concerns over shortages of vaccines for seasonal influenza, and possibly pandemic flu, have fueled a resurgence of interest in cell culture manufacturing for vaccines. Many live and inactivated virus vaccines, such as those for polio, rubella, varicella and rabies, are currently produced in mammalian cells. Novartis (Marburg, Germany), which acquired Chiron in April 2006, operates an egg-based influenza vaccine manufacturing operation in Europe, but earlier this year submitted a product license to European regulators for a cell culturebased vaccine production plant. Novartis expects to receive approval for a flu vaccine produced in mammalian cells by mid-2007. Others in the cell-culture flu vaccine space include GlaxoSmithKline, Sanofi-Pasteur and Baxter.

Novartis’ VP of manufacturing, Norbert Klein, Ph.D. is convinced that the future of flu vaccines lies in cell culture, a conviction shared by the EU’s regulatory agency, EMEA. One may wonder why cell culture does not already dominate the flu vaccine market. Klein says the explanation lies not with cell culture but in the composition of influenza vaccine and influenza’s biology. Flu vaccines incorporate three virus subtypes which change every year. “Some grow very well in eggs, some do not. The virus’ behavior in cell culture as well as in eggs is unpredictable.”

Novartis uses Madin-Darby Canine Kidney (MDCK) cells to grow influenza virus, which readily infects this workhorse cell line. The challenge is to develop a process versatile enough to cope with new strains – influenza’s “moving targets.” Since these virus types change every year, there is practically no room for error.

Innovation is also being seen in vaccine formulation, format and delivery. MedImmune (Gaithersburg, Md.), for example, ships its FluMist live attenuated influenza vaccine frozen, but the company has recently submitted a supplemental biologics license application (sBLA) to FDA for a stabilized liquid formulation stable at refrigerator temperatures. This difference of 30-40 degrees will facilitate handling and may improve acceptance of FluMist’s novel delivery system among physicians and patients.

The principal differences between the frozen and refrigerated formulations are a proprietary component that enhance temperature stability, plus an additional ultracentrifugation purification step. For now, FluMist is produced in eggs, as are all US-licensed flu vaccines, but MedImmune has recently received a contract from the U.S. Dept. of Health and Human Services to accelerate its efforts at developing an influenza product from mammalian cells.

Cancer Vaccines Show Promise Fifteen cancer vaccines, with the potential for curing this dread disease or inducing lengthy remissions, are currently in Phase III testing. The National Cancer Institute funds or sponsors most of these studies, but eventual approvals could put companies like Cell Genesys, Medarex, Genitope and Antigenics on the map. Kalorama Information (New York, N.Y.) calls cancer vaccines “one of the most highly anticipated pharmaceutical breakthroughs of the new millennium.” In The Emerging Cancer Market, a report issued in November 2006, Kalorama estimated potential demand for cancer vaccines at $481 million in 2007, and predicts that prospects for growth to $8 billion by 2012 is “very real.” Kalorama bases its optimism on the possible 2008 approval of up to five cancer vaccines with $700 million in anticipated revenue. Another eight licenses by 2012 could position cancer vaccines among the leaders in oncology therapeutics. Analyst Melissa Elder, who authored the report, says that despite their great promise, cancer vaccines will have to fight entrenched chemotherapy and radiation oncology products. Besides demonstrating efficacy and cost-effectiveness, vaccine manufacturers must show superiority to current treatments before insurers will reimburse their costs. Getting personal One type of cancer vaccine uses material from a patient’s own tumors to induce the immune response – the epitome of personalized medicine. By far the most serious problem facing these drugs is their likely astronomical cost. The thought of a GMP facility concurrently manufacturing thousands of individual “products” is unfathomable, especially for antigenic materials that must be grown or amplified. Platform technologies will relieve some of the burden, but clearly what is needed are compact, automated, rapid, GMP-worthy production devices with fully disposable product contact surfaces. Biovest International, the first vendor to offer a commercial-scale hollow-fiber bioreactor in the mid-1980s, is developing a Phase III vaccine against non-Hodgkin’s lymphoma. In Phase II studies, 95% of treated patients are still alive and 45% remain disease free nine years after treatment. The company has simultaneously been commercializing a compact bioreactor, AutovaxID, which exploits hollow fiber reactors technology installed at hundreds of sites worldwide, to expand patient-derived cells that generate the antigens for the lymphoma vaccine. AutovaxID is suitable for growing any antibody-secreting cell line, including hybridomas and Chinese hamster ovary (CHO) lines. For their personalized cancer vaccine, cells harvested from patients are fused with mouse hybridoma cells, and expanded in the AutovaxID instrument. The device is also used at the Biovest National Cell Culture Center to produce monoclonal antibodies for academic research. AutovaxID consists of two components: the instrumentation and an enclosed, disposable bioreactor (cultureware). The disposable component is completely enclosed in a plexiglass housing that slides into, and snaps in place within the instrument. Following an automated fill-flush operation, inoculation and secondary pH calibration, the instrument operates unattended from that point on. Production yields vary considerably, but the company routinely makes up to 5 grams of protein in 45 days.