Since its founding in 1981, Genzyme Corp. has prided itself on a strong base of innovation (Box, below). The company now has over 70 facilities operating in over 30 countries, including 17 manufacturing plants and nine genetic testing laboratories, and a technology platform that encompasses protein, cell and gene therapies, monoclonal and polyclonal antibodies, genetics, small molecules, biomaterials, therapeutic polymers and rapid diagnostics (Box, below).
Approved for commercial production by the FDA in 1996, Genzyme Corp.’s Allston Landing Facility in Allston, Mass. is one of the largest cell-culture manufacturing plants in the world. The 185,000-sq.-ft. plant, which produces orphan drugs for rare diseases, employs 400 people. It first began producing Cerezyme, an enzyme replacement therapy for Type 1 Gaucher disease, in 1996. Today, Fabrazyme, an enzyme replacement therapy for Fabry disease, and Myozyme, an enzyme replacement therapy for Pompe disease, are also made at the plant.
Increasing manufacturing capacity by 50% in 2005 allowed for additional simultaneous cell-culture and purification, and the plant currently boasts a 12,000-L perfusion capability. Although the facility has no formal Six Sigma or Lean initiatives under way, its management is committed to a continuous improvement program that uses tools from both methodologies, selecting those that are most relevant to each specific stage of manufacturing and quality control (QC).
Utilizing detailed process mapping and cycle time analysis, manufacturing professionals at the plant have eliminated nonproductive steps, says Frank Byron, vice president and site manager at Allston. Meanwhile, both manufacturing and QC teams analyze first-pass yields and corrective and preventive actions (CAPA) to prevent any deviations. The facility is now investigating rapid microbiological analysis, Byron says, as well as implementation of a new finite scheduling system, which he believes will optimize and integrate its production and maintenance scheduling technologies.
At Allston, employee-directed cross-functional improvement teams use tools that include statistical process control charting, statistical analysis on process variables and a process data historian that captures key process variables, to analyze process trends. Trend data are distributed to all via an interactive intranet system, Byron says, and plans call for an integrated process monitoring system to be installed at the plant in the future.
QC Walks the Line
QC is key to operational excellence at the Allston facility, and the typical validation run for any product at the facility averages 100 days.
Media preparation, the first step in the manufacturing process, results in nutrient solutions that help support cell culture operations. In order to keep cultures productive, 50,000 liters of media must be supplied weekly. While sterility is of utmost importance at each step in the manufacturing process, it is particularly critical during media preparation, since any contamination may interfere with protein synthesis. Genzyme uses automated processes to manufacture over 30 varying forms of buffer solutions; raw materials are dissolved with water in mixing vessels and media conditions then adjusted appropriately. “Critical data points such as temperature, density and overall process quality are monitored continually,” explains Andrew Croteau, associate director of QC microbiology.
The importance of process safety and process control cannot be overemphasized during production, says Croteau. “From the time you design a facility like this, you have to think about safety mechanisms, and the varying safety hazards here.”
For instance, there are very strict safety programs such as HAZCOM training to avoid accidents involving high pressure systems, hot-water injection fluids running at 80° C or higher, caustics or bacteria infiltration. “Everyone here, from scientists to manufacturing administrators, is required to go through multiple hazard and technology programs to update our precautions continually,” he points out.
Using CFD to get inside a bioreactor
Genzyme’s bioreactors have been designed and validated for multipurpose functionality; this is supported by diligent monitoring and QC.
Nearly one million gallons of bioprocess fluid and growing cells flow through Allston’s six-gallon stainless bioreactors each year. To better understand what is going on within these reactors, Genzyme plans to use computational fluid dynamics (CFD) to model the cell-culture process.
Multipurpose functionality guides manufacturing, and the bioreactors have been designed and validated so that they can be used to manufacture different materials. A delicate system of checks and balances is at work, Croteau explains. “The bioreactors are dedicated at any one time to each particular drug, however, upon validation we are able to make other products in the same bioreactor,” he says. “We do have to prove that we are able to clean the equipment as required between runs, but we have a very concise system of operations that allows for multipurpose functionality.”
In the case of the enzyme replacement therapy for Fabry disease, Fabrazyme, the CHO (Chinese hamster ovary) cell line was chosen because it is less vulnerable to contamination by human viruses than cells of human origin. Production begins with genetic modification of a host cell to produce -GAL from human DNA. The human gene is isolated, spliced into a bacterial plasmid, and inserted into the CHO host cell in the seed lab. Genetically modified hamster ovary cells are used for Cerezyme and Myozyme as well, and stored in this facility.
Under carefully controlled conditions, the cell grows in a liquid medium of about 50 different nutrients such as sugar, salts and amino acids. After weeks of growth, the cell cultures are finally harvested for purification. Harvested material passes through six 200-liter flash chromatography columns, including the enzymatic modification of the purified product. The final formulation of the solution must be adjusted accordingly to contain exact active concentrations. Once again, rigorous testing is administered and samples gathered every 24 hours to prevent virus particles from entering the final manufacturing phases.
The purification station operates continuously, with operators examining product samples every 24 hours to evaluate cells and ensure the overall health of the culture. With QC testing driving Genzyme facilities, cultures are sent from the bioreactors to test enzyme population production. Throughout the production process, the cells remain continuously in suspension, using up to 2,500 liters of medium per day.
The purified cell is then stabilized with excipients and undergoes double-sterile filtration before it is distributed into vials under aseptic conditions. A typical sterile fill operation takes at least a week of preparation, and more than 20,000 vials are filled in a six-hour period. The vials are lyophilized in a 150-liter filter-dryer to enhance stability and storage. Once lyophilization is completed, each vial must undergo a series of quality-control tests before being shipped from the facility to the patient. “Once the vials have been filled, they must be analyzed by QC for viral assays and microplasmas, and many certificates of analysis must be completed to ensure the drug is safe for our patients,” explains Croteau.
For sterility assurance in the Class 100 fill/finish area, Genzyme relies on adherence to established and validated operating systems. Continuous operator training is critical, as are rigorous standards for cleaning and disinfection of the area, Croteau says.
To verify the facility’s control over its processes and equipment, training sessions are held regularly, including periodic mock fills of product utilizing bacterial growth media. Operators are also specifically trained how to gown for entry to the area. Each operator must be swept for bacteria three times before being allowed in the fill/finish suite.
Genzyme also utilizes HEPA-filtered air to ensure that the manufacturing environment is clean and maintained at a positive pressure to prevent ingress of bacterial contamination. Facility walls are constructed of non-shedding materials that can withstand the application of antibacterial solutions, and to prevent particulate shedding, all equipment is constructed of stainless steel.
Improve, Innovate, Advance
Facility managers at Allston view QC as the gatekeeper for each manufacturing stage. Samples from each manufacturing stage are gathered and examined before further work can be performed. However, QC and manufacturing methods are not viewed as being set in stone. Genzyme evaluates any new technologies being incorporated into operations to ensure they will produce consistent and reliable products.
On the process-control side, the facility’s implementation of fieldbus technology has helped the facility in several ways (Pharmaceutical Manufacturing, September and October, 2006, pp. 41 and 32). Fieldbus not only saved space by reducing wiring requirements, Byron says, but the system has also provided operators with much more information about processes, since they now see “flags” when problems occur with a specific instrument or device.
The company incorporates rigorous validation and Design of Experiments methods into its drug development and product launch programs, providing a basis for future understanding of process as well as a system of monitoring and controlling process change. The company’s Quality organization utilizes development program data to determine critical process points in manufacturing and thus establish sampling methodologies and an effective monitoring program.
To ensure that processes are adequately controlled and reproducible, process analytical data are routinely revalidated and trended. New technologies and better understanding of its manufacturing operations have provided visible bottom line benefits, says Byron. Within the past five years, Allston has reduced by 20 days the QC cycle times for testing and releasing some of its products, but Byron says deviations have dropped by 10% and the cycle time required for their resolution has been cut by 60%. On-time starts for purification are now roughly 99%, Byron says, attributing that to scheduling improvements driven by process mapping.
“We actively evaluate our manufacturing process for consistency and seek opportunities to improve our understanding of our operations,” says Byron. This desire for continuous improvement has been critical to the facility’s success so far.
PM: Do you think public policy stifles drug development? Which country seems to have the most efficient policies in place?
HT: As one of the world’s largest biotech companies, our major responsibility is to the patient. Policies for drug development are necessary to get these drugs and technologies to the patient safely. France has the fastest processes and most efficient policies in place from drug development to distribution.
PM: What impact will patent reform legislation have on biotech?
HT: Patents are very important in drug development and any new reforms will have an impact. It is hard to predict what will come from patent reform legislation, but with a 99.9% failure rate, it often takes years to reach the point of considering patents. There are many hurdles throughout the process, and when all the pieces of the puzzle finally come together, the patent process must be an easy and quick one. Premature licensing of patents can create enormous hurdles for technologies to come together. Patents are a double-edged sword. They differ for all types of technology and with a trial-and-error time span of 6 months to 15 years, patents must allow for obstacles.
PM: What is the most profitable area for Genzyme?
HT: The only thing we care about is our ability to finance continuous innovation and to do this effectively. As a corporation, it is our responsibility to try and bring new ideas together to help patients. We find that the profitability of our company cycles up and down. Mature products, such as Cerezyme for Gaucher’s disease, drive our funding. The investment requirements are high, but that fact makes sense to our investors. If we don’t innovate, we lose to our competitors. If we don’t bring about new therapies, our patients lose.
PM: What is the next frontier for biotech?
HT: Many academics and start-up companies have great ideas for new innovations. We are always looking for the latest and greatest in therapies. I think the next big thing will be gene therapy. Genzyme has conducted clinical work in gene therapy for the last 20 years. We are looking into gene therapies applicable to lysosomal storage disease, cardiovascular disease, Parkinson’s disease and age-related macular degeneration. I am very pleased with our progress so far.
PM: What can we expect to see in Genzyme’s pipeline in the future?
HT: We are at a very exciting point in our clinical trials of many therapies. We hope to begin Phase III clinical trials for multiple sclerosis therapies this year. Last week we had a 75% success rate in treatment therapies and, while they are still roughly four years from being released, we are very excited. We are in the late stages of clinical trials for therapies in genetic diseases, kidney disease, oncology, orthopedics, transplant and diagnostic testing. We also have development programs in immune and infectious disease.
PM: What will Genzyme’s next blockbuster be?
RG: We hope to have a lot of them, but striving for blockbusters is not really our philosophy. If you’re focused on the patient, sometimes you get lucky and sometimes you don’t. Drugs are modest. Gene therapy will not be a blockbuster. We are working in that area because of our commitment to the patient; it may be the best approach for some things but it is also a bit risky for others.
Today, there is huge enthusiasm for stem cell research and regenerative medicine, but I hope everyone understands just how far away some of that is from becoming commercial. For example, we can grow cartilage, we can grow bone and we can replace skin. But have we grown a fully functional digit? Not yet.
We do have a huge optimism about our therapies and their abilities to help a lot of our patients. In R&D we think about medical need and underserved populations, and hope the rest will come from that.
PM: How are new programs initiated? Do patients come to you and how do you assess the needs of these patients?
RG: We find most of our contacts through meetings and academic conferences. Someone will come to you with a great idea and many of our programs start in this manner, with the exception of the recombinant enzyme. We read a lot of papers, and academia is a huge driver of innovation. The money NIH invests in biomedical research is extremely important. We are huge supporters of NIH funding here at Genzyme.
PM: We hear of bio companies having a very challenging time trying to recruit skilled professionals. What should the government and large bio and pharma companies do to help resolve this problem?
RG: It is difficult to recruit trained professionals. In Boston in particular, we are competing with large companies such as Merck and AstraZeneca that have very fine research organizations. Genzyme has one of the lowest turnover rates, which is good for us, but I don’t disagree that it is tougher than it used to be. For some of the smaller, newer bio and pharma companies, recruiting professionals is extremely difficult due to job security in a high-risk environment. Genzyme offers internship programs for college students, which helps both us and them gauge their potential success with our company.
PM: Which are the most cost-effective drugs and innovations?
RG: If gene therapies work, they will have the best cost-efficiency ratio. Cell therapies are going to be extremely difficult to manufacture and scale-up. Vaccines are relatively inexpensive, especially for underdeveloped countries, so we hope to have vaccines and gene therapies on the same level.
PM: Is anything stunting innovation? Intellectual property restrictions, for example?
RG: By and large, to develop a complex drug, the number of patents required is extreme. Some of the layers of patents you end up paying for seem almost superfluous. You want to award innovation to core discovery. However, without the intellectual property system there would be no companies.