Isolating the Leadership Gene

The decision to move from Phase II to Phase III trials is a key moment for a potential medicine. Phase III trials cost up to several hundred million dollars, so you only go forward if you've got something worth investing in. At the same time, if a drug turns out to be effective, a month or two's delay, even if buried back in the process, becomes millions in lost sales. "This was a critical program," says Mullen, "at a critical stage, and we weren't making the progress we should have been."

One Friday, Mullen was at yet another meeting with the Amevive group. They'd brought the data sheets for each patient enrolled in the trials. "I said, 'Give me the data sheets,' " Mullen recalls. "They handed me one sheet, for one patient. I said, 'No, give them all to me. Right now.' "

Over the weekend, Mullen pored over the data, patient by patient, making histograms of Amevive's performance. "I discovered things about the drug that none of these doctors knew," he says, "because they hadn't taken the time to look."

This turn of events is no secret at Biogen. Indeed, for Mullen, the story is an anti-leadership parable: Doing the data analysis himself was the right solution to a problem that never should have existed. It was almost an intellectual temper tantrum. In Mullen's world, ego is a sorry substitute for judgment, and judgment begins with data. "That group had the wrong values for this company," Mullen says. "FYI, they no longer work at Biogen."

Medical Marvels (II) Chinese-hamster ovary (CHO) cells are happy and productive in a brown broth of nutrients held at the same warmth as the human body. In order to make Avonex, Biogen's MS drug, copies of a human gene are inserted into CHO cells. The gene is the one that tells a cell how to assemble interferon beta 1-A, a substance that everyone's body makes in small amounts. Each CHO cell doesn't get just one copy of the interferon-making gene; it gets hundreds of copies. The CHO cells crank out interferon in massive amounts. They also reproduce themselves, and the new cells also crank out interferon in massive amounts.

The process starts with a vial of CHO cells the size of a test tube. As the cells reproduce, they get larger and larger containers in which to live until, after four weeks, they end up in a 2,000-liter tank. What is the result of those CHO cells turning out all that interferon? When the medicine that the CHO cells have made is purified out of the soup, what is left is one-tenth of 1% of the original volume -- a container of concentrated Avonex the size of a 2-liter bottle of Coke.

Faster Is Better

At 44, Gunther Winkler is a veteran of Biogen, his tenure stretching back to 1988, when Biogen was burning through money. In those days, it thought of itself as a "research boutique" -- with 250 employees working on 105 development projects. Winkler remembers his first impression of the place. He was staying at the Marriott hotel across the street, and the night before his job interviews, he couldn't sleep. Several times he looked out the window. "Every time, there were people in the lab, all night long," he says. "I said to my wife, 'That's the kind of place I want to work.' People put their heart and soul into discovering drugs."

But commercializing drugs requires more than heart and soul. It also requires discipline and accountability. For almost half of his time at Biogen, Winkler headed clinical operations, the group that runs human trials. For much of that time, Mullen ran manufacturing. (Even before a drug is approved, companies have to make it in order to test it.) "Biotech manufacturing is very different from classical drug manufacturing," says Winkler. "You have large tanks with thousands of liters of broth in them, and the cells swim around, and they have their own temperaments, of course. And it's easy for them to become contaminated. We'd had a string of 'bad luck' with our manufacturing. When Mullen took over, he figured out how to push the bad luck out. He turned manufacturing into a science here."

Winkler is now the program executive in charge of Amevive, the psoriasis drug awaiting approval by the FDA. Once human trials are finished and a company thinks it has a viable, valuable, and safe product, the company has to submit an application to the FDA to market and sell that drug. The application is really a story, an argument, about the drug, its value, its effectiveness, and its side effects, along with the test data to support the story. Applications run to hundreds of thousands of pages.

The fact that Amevive is six months into the lengthy review process at the start of 2002 is itself another bit of Biogen lore -- a further chapter to the saga of the weekend data sheets. In the spring of 2000, Winkler sat down with Mullen to talk about planning Amevive's FDA application. Human trials were nine months from completion. "It was a routine meeting, and I presented a timeline for completing the application. It was for six months, 180 days. Jim said, 'Gunther, I have a challenge for you. I want you to do it in 90 days.' Frankly, I was speechless. I took it to my team, and they were as speechless as I was. We were simply not sure it could be done."