Biology calls shots in making of drug

Second of three parts

Jim Thomas has spent 20 years coaxing hamster cells to make proteins that improve people's lives.

The work has taught him and his crew at Immunex some serious lessons. He has learned how biotech drugs are fundamentally different from traditional pills and how much more difficult they are to mass-produce.

The reasons for that come down mostly to chemistry vs. biology.

Thomas and fellow scientists at the Seattle company can't rely on mixing chemicals in a flask to produce a drug, as pharmaceutical companies do. To make a drug like Enbrel, they use biological techniques that are boundlessly complicated: They try to manipulate living cells into making drugs.

"The cell really has all the machinery," Thomas said. "We're getting the cell to do all of this for us."

The process is extraordinarily complicated, time-consuming and expensive, which means patients may have to wait months or years — and pay more — for breakthrough drugs while biotech companies struggle to make them efficiently.

In the past, making drugs was relatively easy. Companies like Merck and Pfizer have spent decades doing research that had little to do with cells. They test relatively simple chemical compounds on animals, and test the most promising ones in humans.

Once the drugs pass regulatory scrutiny, they sell them.

Tomorrow in the Times The Enbrel shortage and how it affects patients.

In principle, "small molecule" drugs work like a shotgun. They come in tablets or capsules that are absorbed in the bloodstream, causing chemical reactions that spread through the entire body, not just the sick parts.

When scientists find a chemical that seems to treat disease in animals, they make large batches by heating, cooling, stirring and mixing in other chemicals in a tightly controlled environment. Once mixed with inactive ingredients and tested for quality, the manufacturing process can be done in a few weeks.

Biotech drugs are completely different.

Most use copies of natural proteins or antibodies to home in on a specific target in the body. Treatments are usually injected because proteins are too fragile to withstand digestion.

The work that led to Enbrel, Immunex's popular drug for rheumatoid arthritis, started when scientists cloned a key gene sequence in the 1980s.

Then they spent more than a decade figuring out how to turn the sequence into proteins.

The final step was formulating those proteins into Enbrel and making huge quantities of it.

To make Enbrel and other biotech drugs in large batches, Thomas and his crew have to follow an intricate process. It starts by injecting the gene sequence into host cells, usually cloned hamster cells because they are safe and mimic how proteins are made in humans.

Making a traditional pill 1. Lab scientists isolate a chemical that appears to be effective at treating diseases in animal models. 2. Scientists synthesize larger amounts of the chemical by mixing it with other chemicals, heating, cooling, adjusting the acidity or other variables until they have an active ingredient. 3. The active chemical is isolated. It is combined in a bulk powder with dry, inactive ingredients, such as gums that bind it together, diluting agents, lubricants, coloring and flavoring agents. The active and inactive ingredients are compressed together into a tablet form. 4. Tests are done for quality and purity. 5. The drug is labeled and tested again for quality and purity. It is bottled and ready to be shipped to the patient.

The hamster cells are delicate but they survive inside stainless-steel vats filled with a nutrient broth.

That nutrient mixture carries all the vitamins, minerals and amino acids a cell needs to survive as it would inside a body.

As long as scientists keep cells alive with a perfect balance of temperature, oxygen, acidity and other variables, they can coax them to make proteins.

Keeping that perfect balance is extremely difficult, scientists say. The temperature, for example, can vary by no more than one degree Celsius.

The acidity is even more critical: Cells produce carbon dioxide, which constantly changes the pH levels, and if those levels swing even a tiny fraction, cells can die.

After several weeks, the proteins are filtered away from unwanted by-products. The end product is tested for weeks to ensure no misshapen proteins made it through screens and that no unwanted bacteria, viruses or carbohydrates hitchhiked along the way.

The process is so intricate that Jeff Richardson, spokesman for California biotech giant Amgen, likened it to "making a snowflake the same (shape) every time."

The techniques have proved difficult for Immunex. Seattle's leading biotech company hasn't been able to keep up with demand for its drug, and it has spent months tweaking processes to squeeze out at least 10 percent more from its factories.

Compare that with the typical pill, which scientists say is relatively easy to manufacture.

"The hard part isn't making a small-molecule drug, it's figuring out which chemicals might work on diseases," said Rick Lapointe, vice president of operations at Hollister-Stier in Spokane, which does contract manufacturing for biotech and conventional drugs.

Scientists also have to consider what can go wrong with biotech manufacturing. Viruses or bacteria can contaminate the batch at any step. Proteins can be broken down by subtle environmental changes.

And just when scientists think they've created the perfect environment for cells to grow in a 500-liter vat, they find it doesn't work in a 1,000-liter vat or a 500-liter vat that's shaped slightly differently.

They don't know why.

Solving such a mystery isn't easy when every step is a tightly guarded secret within each company.

Partly because of the culture of secrecy, few people in the biotech industry understand the manufacturing process. High costs and companies' reluctance to part with recipes helped drive many manufacturers out of business in the early 1990s. Now there's a worldwide shortage of factory capacity.

Dave Vetterlein, head of manufacturing at Bothell-based Icos, said there are fewer than two-dozen places in the world, including Immunex and Icos, that can produce large quantities of biotech drugs.

Even the people who are among the best at it, such as Vetterlein and Thomas, know their limitations.

"These are living organisms we're dealing with," Vetterlein said. "Understanding them is like asking how we understand life, which has evolved for billions of years. We're just scratching the surface."

The process is so baffling that scientists have a word for cell behavior. They call cells "happy" if they are staying alive and not producing too many misshapen proteins or carbohydrates.

Cynthia Robbins-Roth, founder of BioVenture Consultants and an early scientist at California biotech-manufacturer Genentech, said the process is a developing art.

She remembers a time at Genentech when many cells died in the summer and scientists couldn't understand why. Eventually, they figured out the city added chlorine to the water, and traces of it had slipped through purifiers.

"It's not easy when you have a biological organism making a drug," Robbins-Roth said. "Inherently, it's not as easy to control. Mother Nature throws lots of little wrenches in the way."

Vetterlein said that any improvements in the process will take years, but that it will happen because the field is advancing, though there's still a lot to learn.

"Basic drug chemistry has been around for hundreds of years, and the laws that govern those chemical reactions are pretty well understood," Vetterlein said.

"With a biotech drug, you're basically in control of a living system. It's kind of like how you keep a rabbit alive. It's not like you're in total control, it's partly the rabbit."

Luke Timmerman can be reached at 206-515-5644 or ltimmerman@seattletimes.com.