The Antidote: Inside the World of New Pharma (14 page)

Thomson managed Vertex’s drug discovery machinery in Cambridge even as he built it; promoting from within where possible, shifting around responsibilities, “picturing problems in a more structured manner and making that a template for solutions,” he recalls. As he tried to arrange the pieces to give fabric to an organization that was moving on two tracks—Novartis, and everything else—he also continued to drive the Lilly collaboration, endeavoring to identify and convince key managers in Indianapolis that VX-950 was a developable compound. “We were in their faces, and that made them nervous,” he says. Thomson stopped smoking at work, striving to be less wild man and rogue, more exemplary.

Sato assigned Murcko overall responsibility for bringing Vertex and Aurora together and assigned Tung to teach Aurora to hunt for drugs. Each welcomed the assignment. Murcko wanted to ensure that Vertex kept expanding its possibilities and didn’t bog down in one approach to drug discovery, even one as powerful as design. Tung remained skeptical of Vertex’s boldest claims for design and was eager to test his own theories outside the mother ship of the Cambridge labs.

Tung and Murcko alternated weeks in San Diego while Aurora’s senior management team, absorbing the impact of the merger, reshuffled. Landing on top after the shake-up was biologist Paul Negulescu, formerly head of research, who now ran the site for Vertex. Negulescu was even-tempered, deft, strong willed, enthusiastic, and patient, a true believer in Aurora’s high-throughput screening who had joined the company at age twenty-seven fresh from a postdoc in biophysics and immunology. One of Aurora’s first five employees, he’d been hired as director of biology, “which meant director of myself,” he recalls. Now he
resolved to “stay through it, make the acquisition a success” by helping make the site fully capable: bringing in PK, pharmacology, and medicinal chemistry to help improve and advance the hits emerging from Aurora’s “cell-based stuff.”

Negulescu reported to Murcko. “We had a lot of projects, most of them very, very early, and Mark guided us through the culling process,” he recalls. “The one thing I remember about all of those interactions, whatever the purpose of the visit or the topic, there was always the sense that we would make something good happen. I think that’s part of Mark’s genius. He’s a glass-is-half-full person, and you shouldn’t underestimate how important that was. The support, and confidence he gave us just by being here and saying ‘I believe in you guys’ was probably as significant as anything else.”

Cystic fibrosis was an ion-channel project, one of several types of screens Aurora had developed. After the gene for CF was identified in 1989 by a group co-led by Francis Collins, speculation rose about what it did in the body. Researchers discovered that it encoded for a protein that they chose to name—still with no insight about its function—cystic fibrosis transmembrane conductance regulator (CFTR): literally, the pore-forming protein on certain cell surfaces that lets through charged atoms and molecules and that we know about only because it is broken in people with cystic fibrosis. Soon after, they determined that CFTR channeled chloride ions and water, explaining, ex post facto, how for centuries folk healers diagnosed the disease by licking babies to see if they tasted salty.

Knowing what CFTR did made it only more daunting as a target. Hedging its strategy, Aurora proposed in its initial agreement with the foundation to pursue alternative pathways as well. But Beall and his medical staff pressed the company hard to focus on the disease at its heart. “I credit Bob with that,” Negulescu says. “He was going around looking for companies not just to work on CF but on CFTR—
for
CF. He was frustrated by the fact that we knew that CFTR was the defective protein, and we knew that it was a channel, and we knew what was wrong with it, but there was nobody looking at how to fix it. They were looking at alternative channels that one could try to activate to control the symptoms,
say, with inflammatory responses. Or how does one interrupt the bacterial environment in the CF lung? Everything but the real issue.”

Passion is an underrated virtue in drug research. Big Pharma R&D leaders, forever imposing processes and metrics, believe that strategy ultimately prevails; pick an approach, stick to it, be meticulous, you’ll get there. But Beall’s deep conviction that only a drug that hit CFTR could transform the lives of people with cystic fibrosis had inspired a growing faith at Aurora, if not yet at Vertex.

Negulescu flew to Cambridge to lay out the goals of the CF program before the scientific advisory board. Murcko, anticipating resistance to Aurora’s approach, cautioned him not to expect a warm reception. What Aurora would do, Negulescu explained, was find compounds to correct each of the two main types of defects. In some cases, there’s sufficient CFTR at the cell surface, but the channels don’t remain open long enough to let the ions and water pass through—so-called gating mutations. In many more patients, the problem is that not enough protein gets to the cell surface, because the protein is improperly folded. When Negulescu said that Aurora was screening for molecules to activate the protein regions responsible for both problems, an advisor muttered, “Fantasy science.” Boger and Sato, sitting in back, said little.

In April microbiologist Eric Olson took over the CF project. He was unlike Kwong and others who, inspired by the company, chose to join Vertex because that’s where they thought they could do their best work: in Olson’s case, Vertex and CF seemed to come together to select him. A lean, blond, soft-spoken Minnesotan, Olson had worked in antibiotics for sixteen years at Upjohn and Warner-Lambert, recently acquired by Pfizer. He’d become interested in CF through a colleague whose daughter had the disease, and they’d collaborated against pseudomonas, the most common bacteria to attack the CF lung. Living in Ann Arbor, looking a year earlier for a job, he’d interviewed at Vertex for a position as program director, spending an hour with Sato and coming away impressed. Meanwhile, he also interviewed with Negulescu at Aurora, before the merger. He got an offer from Lilly, but heard that Lilly might be soon getting out of anti-infectives.

“I almost just about didn’t join Aurora because even though they
could set up screens, there was no way they could ever make a drug,” Olson recalls. “But I got wind when I was out there that they were in deep discussions to solve that problem. They said, ‘Don’t worry about that.’ I did worry about it, but I really trusted Paul. A couple of weeks later, I was in the airport, going to Iowa, when the thing crosses the news: ‘Aurora bought by Vertex.’ I was so glad, since I’d already gone to Vertex. They each had a piece that I felt could make a big difference.”

Olson’s group had developed two related approaches to measure CFTR activity. Both employed high-throughput “patch clamp” assays in which swatches of cell membrane are isolated using a micropipette tip, and then hooked up to microelectrodes to gauge the opening and closing rates of individual channels. Screening for molecules to activate CFTR with gating mutations—so-called potentiators—technicians looked for changes in the electrical flow over a two-minute period. For those that might fix the folding problem and get the protein to the surface—correctors—they ran the same experiment but overnight, giving time to allow more protein to activate and get to the membrane. After sifting through tens of thousands of compounds, in June Aurora recorded its first validated hit for a potentiator. The molecule wasn’t potent or specific; much less was it known to be effective, safe, or easy to formulate and manufacture. But it showed that a small molecule could enhance the functioning of CFTR.

Olson, Negulescu, and the CF team regarded the hit as an important moment, but they neither expected, nor found, real corresponding enthusiasm in Cambridge. Having a hit meant the true start of a project, and Tung and others did what needed doing to move the molecule along. But as seasoned drug hunters they remained skeptical. “They came out and set up all the stuff that had to be set up, but it wasn’t clear what the feeling in Cambridge was,” Olson recalls. “My sense was, ‘Look, the foundation’s covering most of the cost; you guys want to keep doing it, go ahead.’ In the meantime, they were trying to advance all these clinical programs anyway; they weren’t ready to start a bunch of new projects.

“Let’s put it this way,” he says. “Nobody stopped it.”

The first annual Liver Meeting® at the John B. Hynes Veterans Memorial Convention Center auditorium in the Back Bay during the first week of November was an attempt by both its sponsors and the city to rename, rebrand, and trademark the half-century-old yearly get-together of the American Association for the Study of Liver Diseases (AASLD), where until the last few years the reigning topic was cirrhosis. They hoped to make the meeting—and the expensively refurbished three-tier conference hall—an important venue in the business of biomedicine. It wasn’t Vertex-Lilly’s VX-950, but another company’s molecule, that stole the show.

In four papers describing the discovery, safety, and early antiviral activity of a small-molecule inhibitor of HCV protease, Boehringer Ingelheim’s BILN-2061 proved beyond a doubt what Vertex, Schering, Roche, Merck, and many other companies had claimed on faith for a decade: that by blocking the protease with a selective inhibitor and making it orally available, you could dramatically reduce the amount of virus in patients. In a group comparison with a placebo, an inactive pill, all ten patients taking the drug demonstrated after two days more than a hundredfold decline of HCV RNA in their blood. When they stopped taking the medicine, the numbers returned.

Whatever the sting of losing an apparent lead in the public race for a cure, Sato, Kwong, and the HCV team were buoyed and influenced far more decisively by Boehringer’s affirmation of the target. They now knew that the first company to bring approved protease inhibitors to the millions of infected and sick people with hepatitis C would usher in a new age of treatment, just as with AIDS. Kwong’s group, still struggling to bioengineer an animal model to resemble the human liver, doubled down. Many, including Kwong herself, also drove “midnight projects”: independent side experiments encouraged and supported by Vertex.

The rumblings Olson heard about Lilly turned out to be true: after several fitful years of on-again, off-again commitments, and the debacle of losing Agouron’s AIDS drug, the company was pulling out of anti-infectives. Vertex’s main champion in Indianapolis was transferred. Lilly began requesting amendments to its licensing agreement that would put more of the onus—and costs—on Vertex for clinical proof of concept.
From Aldrich’s initial hard bargaining to the widespread belief inside Lilly that VX-950 wasn’t druggable, there was little but bad feeling about the program. As Lilly went through its portfolio in the weeks after the Liver Meeting, VX-950 was discontinued.

With protease inhibition validated by Boehringer’s results, Boger might have felt more liberated and relieved to have the molecule back had he, Sato, Alam, and the scientists been better prepared to meet the costs and risks of going it alone. Pharmacology studies and patient trials were expensive enough; VX-950 cost $2.5 million per kilo to make. The compound was extremely tough to work with. Vertex knew it could be improved, though not, specifically, when, how, or how much. The road ahead would be long, dark. “I don’t think we were being overly optimistic in terms of extrapolating how far we could improve the situation both in the formulation and synthesis,” Thomson says. “But they said, ‘No, given where we are at the moment, we want a calculation that says this has profitable margins, without extrapolating any further progress.’ The program didn’t explode, but we had a stalemate on whether and how to proceed. So the lawyers stepped in to find a solution.”

Investors and analysts, examining what appeared to be Lilly’s rejection of the molecule, found little upside in their restructured settlement. In exchange for worldwide rights to compounds identified during the collaboration, Vertex granted Lilly a small royalty on future sales, a gesture acknowledging that a Lilly chemist had made the molecule that now was a Vertex asset and would-be product.