Toward Greener Therapeutic Proteins

THE PHARMACEUTICAL industry is making great strides in developing cleaner and more efficient syntheses—that is, greener chemistry—for manufacturing small-molecule drugs. But how green are manufacturing processes for biopharmaceuticals, such as monoclonal antibodies, peptide hormones, and vaccines? And can those processes be greener?

Those are questions Sa V. Ho and colleagues at Pfizer's Global Biologics unit, based in Chesterfield, Mo., are trying to answer. The researchers have set out to determine the type and amount of resources required and the wastes generated by mammalian-cell-culture and microbial fermentation processes that are used to make therapeutic proteins. Their goal is to promote cost-saving process improvements that are also environmentally friendly. In doing so, they hope to stimulate other pharmaceutical companies to join them in developing metrics that illuminate the degree of greenness in biologics manufacturing processes, Ho says.

The pharmaceutical industry is one of the first major manufacturing sectors to wholeheartedly adopt and put into practice the principles of green chemistry, which encourage the design of chemical products and processes that reduce or eliminate the use and generation of hazardous substances.

In 2005, after more than a decade of progress, several industry leaders together with the American Chemical Society's Green Chemistry Institute established the Pharmaceutical Roundtable-a working group made up of representatives from major pharmaceutical companies. One of the roundtable's goals has been to identify and overcome common drug discovery and process chemistry challenges related to the production of small-molecule drugs (C&EN, July 30, 2007, page 61).

Pfizer's initiative with biologics naturally follows the Pharmaceutical Roundtable's lead, Ho says. The company's green chemistry leader, Peter J. Dunn, who is cochair of the roundtable, suggested during a visit to Pfizer Biologics that the researchers there should consider the biopharmaceuticals analysis. As a project leader, Ho says he took up the challenge "with a strong endorsement from Pfizer's management."

Ho notes that there are few examples of anyone carrying out this type of analysis for therapeutic proteins, and no one he is aware of has publicly disclosed a complete assessment. "When we started this project, it wasn't clear to us what type of analysis should be done—everything was a blank," Ho observes. "There was some guidance from small-molecule production, but there was little in the biologics arena to guide us."

Therapeutic proteins are by far the largest class of biologics, Ho notes, with the market size projected to reach $70 billion per year by 2010. Glycoproteins, primarily in the form of monoclonal antibodies produced by mammalian cell culture, make up 60% of the market, while nonglycosylated proteins produced by microbial fermentation using engineered Escherichia coli or other microbes make up the remainder, he says.

Manufacturing therapeutic proteins using biological systems is relatively environmentally friendly compared with the production of small-molecule drugs and commodity petroleum-derived chemicals. The modified mammalian cells or microbes serve as tiny chemical factories in aqueous systems with few, if any, hazardous chemicals required and low volumes of organic solvents needed, Ho says.

But the downside is that the processes use a lot of water, Ho points out. Batch reactors hold thousands of liters of culture or fermentation broth, he says. Even more water along with consumable processing aids, such as tubing, filters, and chromatography resins, are needed for downstream purification steps to remove reagents, host-cell proteins and DNA, and other impurities. Additional processing is required in some cases to convert the purified therapeutic proteins to their bioactive forms and then to prepare stable drug formulations.

AS A STARTING POINT, Ho and Pfizer colleagues Joseph M. McLaughlin, Andrew C. Espenschied, Robert E. Kottmeier, and James F. Bouressa decided to systematically evaluate various bioprocesses by determining the total amount in kilograms of water, organic solvents, reagents, and consumables used per kilogram of product produced. This per-kilogram value, called the E Factor, is an important metric used by process chemists. It was introduced about 20 years ago by organic chemistry professor Roger A. Sheldon of Delft University of Technology, in the Netherlands.

Because the E Factor considers only the amount and not the nature of the material used or waste formed, it's still not a true measure of the environmental impact of chemical processes. For example, 1 kg of treated water doesn't have the same impact as 1 kg of sodium chloride, 1 kg of a chromium salt, or 1 kg of dichloromethane. But chemists can add a weighting factor for each input or waste to help balance their analyses.

Small-molecule drugs have among the highest E Factors in the chemical industry, with 25–100 kg of material consumed per kg of product, according to Sheldon's analyses (Green Chem. 2007, 9, 1273). For comparison, the petrochemical and bulk chemicals sectors consume 5 kg or less of material inputs per kg of product. Even at a nominal disposal cost of $1.00 per kg, the potential annual savings in waste avoidance are significant. This bottom-line opportunity is a major reason that the pharmaceutical industry is leading the way in adopting green chemistry.

When comparing E Factors across different sectors, there are a couple of caveats, Ho says. Commercial production volumes of drugs are much less than that of basic chemicals, but the difference often is offset by the fact that small-molecule drugs require many synthesis steps and patient safety demands high purity. Sheldon's E Factor values also exclude water to prevent the numbers from becoming unwieldy, whereas the Pfizer analysis includes water because it plays such a critical role in biologics manufacturing.

The analysis team initially looked at data for 19 developmental small-molecule drugs compiled by members of the Pharmaceutical Roundtable. The researchers then turned to biologics, made by Pfizer and other companies, dividing the analysis into manufacturing processes for monoclonal antibodies and for nonglycosylated proteins. For each set, they determined the amount of materials needed for production, including water, pH buffers, nutrients (sugars and trace minerals), and other additives. They also determined the requirements for downstream processing, including water, buffers, inorganic salts, chromatography columns and resins, and solvents for chromatography. All in all, the typical bioprocess has about two dozen inputs.

The 19 small-molecule drug processes generated an average 174 kg of input per kg of product, with about 100 kg of that amount attributed to organic solvents—most of it treated as hazardous waste—and 50 kg of water. Minus the water, these values for the unoptimized processes are in accord with Sheldon's values.

As for biologics, a current large-scale cell-culture process to make a kilogram of monoclonal antibody requires more than 7,600 kg of material, divided up as 7,000 kg of water, 600 kg of inorganic salts and buffers (which end up in the aqueous waste), 8 kg of organic solvents (primarily alcohols), and 4 kg of consumables, Ho relates. A medium-sized protein made in E. coli requires some 15,500 kg of material per kg of product, divided up as 15,000 kg of water, 400 kg of inorganic salts and buffers, up to 100 kg of organic solvents—some of it hazardous waste—and 20 kg of consumables.

The researchers took a further look at what might be possible for streamlined monoclonal antibody production by estimating the amount of material needed for a hypothetical highly optimized process. They found that the amount of material used could be cut in half, from about 7,600 kg to about 3,300 kg, Ho says.

EXCLUDING WATER, E Factors for therapeutic protein manufacturing currently are about five times higher than for small-molecule drugs, according to the Pfizer team's initial analysis. These values are only order-of-magnitude estimates for typical processes, Ho emphasizes, because biomanufacturing can vary widely, especially for processes that employ microbes.

"The take-home message is that water use could be reduced, perhaps significantly, if we pay attention to improving manufacturing technology," Ho says. In the end, the water needs to be treated, and the more water there is, the more it costs to treat. But in some cases, the water could be recycled if there were a strong economic and/or environmental incentive to do so, he adds.

As the analysis continues, Ho and his colleagues will be looking at other aspects of biologics processing in more detail, such as the use of consumables—some of which can be recycled—and disposable equipment. "In small molecules, chemists diligently try to reduce consumables because their use is wasteful and costly," he says. "We are now looking to apply the same focus on the biologics side."

At the same time, the biotech industry is already moving toward a greater use of disposables and, Ho points out, disposable equipment manufacturers are pushing hard in that direction (C&EN, June 4, 2007, page 20). Stainless steel tanks and reactors can be used for a long time, he explains, but you have to prepare and clean them for each use and maintain them for a long time, too. One-use bioreactors—tank-shaped plastic bags with built-in fittings—and other disposable equipment will help reduce water consumption because less water will be needed for cleaning, sterilizing, and other steps, he says.

Using disposables, however, can substantially increase solid waste. "The relative environmental impact of disposables versus the traditional approach will need careful and complete assessment," Ho says.

Now that the Pfizer researchers have some background data, "we can create a framework within which biologics manufacturing can be analyzed from a green technology and sustainability perspective," Ho notes. The team will continue to pursue its analysis, going deeper, for example, to look at energy usage and the development of continuous-flow reactors, nonchromatographic separations, and other production platforms such as cell-free synthesis and genetically engineered plants.