Novartis' stunning new research home in a former candy factory is but one ingredient in a bold new recipe for success that embraces leadership, culture, science, and technology.
By Kevin Davies and John Russell
July 14, 2004 | Paradise is the name of a certain male strip club on the fringe of the MIT campus. It is also a reasonable description for the spectacular, newly renovated home of its well-to-do neighbors on the same block, the Swiss-American pharma giant Novartis.
For 75 years, the New England Confectionery Company (Necco) building has been a hallmark on the Cambridge, Mass., landscape. But in an extreme makeover this spring, the former home of candy hearts was transformed into the global headquarters of the Novartis Institutes for Biomedical Research (NIBR). The first of several hundred researchers moved into the pristine new labs and offices in May.
Novartis' decision to relocate its research hub from Basel to Beantown was a major surprise, given Boston's already crowded biomedical landscape and exorbitant cost of living. Even though several big pharmas have a presence in the region, including AstraZeneca, Merck, Pfizer, and Wyeth, Novartis' arrival goes far beyond validating the region's richness of medical and intellectual talent.
"We needed to draw on talent, and it came down to California and Boston," says Fintan Steele, global vice president for communications. Boston won out because it is "six time zones from Basel instead of nine — but more importantly, it has ready access to Harvard, MIT, biotech, and hospitals," such as the Dana-Farber Cancer Institute.
|NOVARTIS: derived from the Latin "novae artes," meaning "new arts, new skills, new methods, new character."
NOV-AE: new, novel, unusual, extraordinary.
AR-TES: (1) arts; (2) skills, methods; (3) character, method of acting.
Hiring has been a breeze. "It's been amazing," Steele says. "For the past year, we've been hiring at a rate of 1.8 people per business day. We've had over 7,000 applications for 450 positions!" (see "Turning Sugar into Science," May 2004 Bio·IT World
, page 58). He admits to one concern, however: "We didn't want to look like we were stealing people!" Fortuitously, cutbacks at Millennium Pharmaceuticals provided Novartis with "the pick of some of their best researchers." So many, in fact, that NIBR banned display of Millennium bags and T-shirts.
The decision to establish NIBR in Boston signifies a rapidly changing culture at the Swiss pharma that permeates from the very top, in the form of CEO Daniel Vasella, who has run Novartis since the 1996 merger between Sandoz and Ciba-Geigy, and NIBR chief Mark Fishman.
Be More Rational
Two years ago, Vasella boldly decided to create an entirely new kind of research headquarters — one that focused on the underlying mechanisms of disease rather than relying on the traditional model of disease-specific silos. And he made plans to build a state-of-the-art headquarters that puts a premium on interaction and collaboration.
Steele explains that Vasella, inspired by the success of the cancer blockbuster Gleevec, thought, "If we understood mechanisms of disease, we'd have more drugs like Gleevec. That needs basic research in functionally characterizing the genome for the sake of drug development."
The desire for a more rational approach to drug development meant learning by example from academia and biotech. Vasella wanted a leader who was not only a clinician like himself, but also an established genome scientist. "Fishman was the person he pursued most aggressively," Steele says.
But Fishman was no easy catch. It took six months before Vasella persuaded the former chief of cardiology at Massachusetts General Hospital to accept the challenge. Since taking up the post, however, Fishman told The New York Times, "I've never had a moment's doubt. I don't know why — any rational person would!"
Fishman emerged unscathed from a two-month crash course at Harvard Business School. During a talk to business leaders last year, he flashed a New Yorker cartoon of two medieval knights in armor, one confiding: "I've never actually stormed a castle, but I have taken siege management courses."
| THE RIGHT STUFF: Sasha Kamb, the new chief of oncology at Novartis, turns from cancer genes to drugs.
Fishman evidently relishes the task ahead of him. "We are doing no less than creating a new art of pharmaceutical science, founded at the intersection of the genomic and chemical universes," he says. He talks of opening "the next frontier of science" following the completion of the Human Genome Project, and the benefits that will bring to patients and society as a whole. And he speaks of building "a new 'sociology' of science, blending the best of academic, biotechnology, and pharmaceutical research cultures."
"Traditional pharma relies on diseases," Steele adds, but Novartis is breaking the mold. "There's still some disease area focus, but when you look at mechanisms, now you're dealing with pathways, systems biology, basic biochemical understanding. We're building new platforms that support these areas."
The "new art" at NIBR, which also includes a facility in MIT's Technology Square, features a functional genomics group, led by Dahlia Cohen (see "Talent Fuels Drug Pipeline in Swiss Time," Jan. 2003 Bio·IT World, page 64), as well as groups devoted to molecular pathways (including hardcore cell biology and computational approaches) and epigenetics, which together will provide the foundation to characterize new drug targets.
A prominent in vivo assay system will be the zebrafish. Fishman pioneered the use of large-scale mutational screens in zebrafish at Mass General, identifying dozens of genes with profound effects on cardiovascular development, including many with direct counterparts in humans. In a recent paper in Nature Biotechnology, Fishman and colleagues demonstrated the potential of this in vivo assay system. By screening a grid of mutant zebrafish embryos in 96-well plates with a library of some 5,000 small molecules provided by Harvard's Stuart Schreiber, Fishman's team identified two compounds that successfully reversed an aortic blockage mutation, restoring blood flow to the trunk and tail.
For any pharma company, building the drug pipeline of tomorrow requires attention to the R&D infrastructure of today. Drug discovery has experienced "a convergence of concepts and technology approaches," says Alex Matter, the former head of oncology who now runs Novartis' Institute of Tropical Medicine in Singapore. "Everything is based on the genome, and the methodologies can be used successfully in one area or another."
Replacing Matter as global head of oncology is Alexander "Sasha" Kamb. Ten years ago, Kamb was the director of research at Myriad Genetics, where he led the discovery of the BRCA1 breast and ovarian cancer gene. Two years later, he co-founded Arcaris, serving briefly (and miserably, he says) as CEO before Deltagen acquired the company, burned through $150 million, and collapsed. Kamb was considering academic positions when a recruiter asked if he'd be interested in biotech companies. "I said no," Kamb recalls. "She said, 'Are you interested in head of oncology at Novartis?' And I said, 'No, I'm especially not interested in that!'"
Kamb didn't take the interview process too seriously. "I'm not the traditional hire by any means," he says. "A biotech guy from Utah, no Big Pharma experience, never been in academia as a professor. But Mark was looking for some different kinds of creatures for ecological diversity ... I heard he'd seen 100 CVs, and didn't like any of them!"
Sporting a wide background that embraces molecular biology, protein crystallography, and genomics, Kamb is seeking similar diversity as he completes his team of some 250 scientists. The conspicuous lack of drug development experience on his resume didn't faze Fishman or head of pharma development Jörg Reinhardt. Kamb recalls telling Reinhardt, "I never did any drug discovery or development." The heavily accented response was simply: "Don't worry, we understand how to develop drugs."
Indeed, the new blood at NIBR boasts several recruits without direct Big Pharma experience, including physician Seigo Izumo from Beth Israel Deaconess Medical Center (cardiovascular research) and Mass General's En Li (mouse models). By contrast, Scott Biller, formerly with Bristol-Myers Squibb, heads a revamped Drug Discovery Chemistry division that has scrapped the traditional disease-specific silos to provide greater flexibility. Fishman has also set up an internal program office under John Hastewell to review NIBR research programs, and to make specific recommendations for areas to implement or cut back.
|The Gleevec Chronicles
|"Stunning Success," trumpets the headline of the full-page advertisement. "Deadly cancer at 23.
For Kamb, Gleevec may be an impossible act to follow (see "The Gleevec Chronicles,"
right), but he is excited: "I feel lucky, because I'm inheriting the fruits of a great organizer and a visionary in Alex Matter." Among his immediate priorities, he says: "We're going to explore antibodies, we're going to support our bread and butter — small molecules — and we're going to look at, but not invest in, gene therapy. There's viral therapy, vaccines ... a lesson from [the race for] BRCA1 is you can't do everything well."
In some areas, Kamb prefers to take a cautious approach. The future of biology will be much more quantitative, he predicts, but he despises overselling technology. Drug discovery "is more in silico-based than 10 years ago, but design a drug de novo from a structure?" he asks skeptically. "I say don't even bother, it just won't work ... it won't ever work without huge changes in how we evaluate drugs."
Aside from the scientific challenges, Kamb must also cope with managing a division of 250 scientists split almost equally across the Atlantic. Videoconferencing helps to a degree, "but it isn't a great substitute when there's a new person in a new role." Frequent trips to Basel are essential so "they develop confidence in me, [and] I need to understand them." Appreciating the cultural differences between the ever-punctual Swiss and the more laid-back native Californian will take a little longer perhaps.
But Kamb is giddy about his new surroundings, especially as his temporarily misplaced computer has just been located. "I love what they did with the interior," he says. "It's got these lovely high ceilings, you feel like you're connected to all the floors, you see people going up and down in the elevators. The skylight at the top — two thumbs up!" He ticks off the "fishbowl" conference rooms, the white mushroom-shaped pillars a la Frank Lloyd Wright, and the informal meeting areas on each floor, complete with cappuccino bars. Attention to detail extends to the DNA double helix motif in the tiled atrium, which, Kamb notes approvingly, has the correct right-handed orientation.
Three Clicks to Paradise
The road to IT and informatics bliss for Novartis may lead directly through the Paradise Café in the shadow of the Necco building. This fixture of Cambridge nightlife has resisted offers to sell (perhaps due to difficulty in getting permits elsewhere).
"Oh, you've heard about that?" shrugs George Morris, for whom Paradise lost would be paradise found. Novartis' chief operating officer for informatics is lusting for a nearby location on which to build a "state-of-the-art advanced computing center."
For now, Morris is happy with his spartan office in the Necco building, as he sets about provisioning the Cambridge R&D facilities with all things IT and informatics. He's got money to spend, dozens of positions to fill, and a spectacular vision to deliver as laid out by Manuel Peitsch, Novartis CIO for informatics.
"I definitely see Novartis maintaining its leadership in using grid (computing), its ability to adapt and use high-performance computing (HPC) in whatever form that will look like in two or three years," Morris says. "We'll also have some phenomenal applications that will essentially let any individual in the organization ... quickly drill in and be three clicks away from whatever data or information you want. That's Manuel's vision, and he's got a lot of us excited about it."
|Novartis' IT Guy
|George Morris swipes his security card to unlock the door to his office and mutters something about the Swiss obsession for security.
Step one, almost completed, is to develop a plan for provisioning basic IT infrastructure. Step two will be the creation of an HPC environment to deliver research applications and data. So far, there's little IT infrastructure in the Necco building. Sure, there are PCs in the labs, and high-speed connections to other company resources, but Morris clearly relishes the challenges ahead (see "Novartis' IT Guy,"
right). A former scientist at the MIT Center for Cancer Research, Morris spent 14 years at Genetics Institute "to reap the harvest of the biotech boom," including its acquisition by American Home Products (later Wyeth). Peitsch lured him to Novartis from proteomics startup Zycos.
One decision had already been made: to adopt the corporate standards for back-office functions. "Basically all of our file serving, all the standards we use for networking, etc., were copied," Morris explains, noting that "new and different technology" doesn't necessarily improve competitiveness.
Much of the IT infrastructure is provided by corporate IT services, a Novartis business unit located in New Jersey. The services group operates like an IT outsourcer. "We're spending a great deal of time establishing the appropriate service-level agreement between the organizations. Their core infrastructure is Unix, and it's based on the IBM platform."
HPC is where the fun and competitive advantages are. Morris plans to create a state-of-the-art facility. All the big box makers are being evaluated, he says, mentioning Hewlett-Packard and IBM by name.
TRICK OF THE TAIL: Zebrafish offer a window into chemical screening. Novartis' Mark Fishman pioneered the use of mutational screens in zebrafish.
Novartis has already made its mark on grid computing with a grid of 2,700 PCs using United Devices technology. Over the next two years, it plans to expand to about 25,000 nodes, with 5 teraFLOPS (5 trillion floating-point operations per second) of processing power (see "Gridlock Is a Good Thing at Novartis," Oct. 2003 Bio·IT World
, page 10). Peitsch says the grid's computing power enables scientists to "carry out in silico
modeling experiments that weren't possible a year ago." Proof of principle came in a 2003 paper describing the virtual screening of a 3-D database that yielded a potent CK2 kinase inhibitor by high-throughput docking.
"There's an in silico sciences group Manuel has created [in Vienna] to serve as the [grid's] gatekeepers," Morris says. "We will be identifying and prioritizing which applications get onto the grid." Demand for grid time is currently greatest from "classic" bioinformatics and cheminformatics applications, and is rising fast.
Linux will also play a major role in the new HPC environment. Novartis, like other big pharmas, looks set to trust Linux in the lab, if not yet with back-office financials. RedHat and SCO are the leading contenders, Morris says, but no choice has been made. He is also carefully monitoring SCO's legal challenges to the open Linux community. "You have to be vigilant," Morris acknowledges. "SCO code is everywhere. We're definitely cautious ... We work closely with all the major [computer] vendors whose Linux we might use, and we discuss risk-mitigation approaches we can put together. Pricing always factors in, too."
Storage is another critical piece to the Novartis IT puzzle. It currently uses a patchwork of vendors and architectures including storage-area networks (SANs) and network-attached storage (NAS). "The usual list of suspects — IBM, EMC, Network Appliance — is being evaluated. We'll use the clean-sheet approach," Morris emphasizes. Oracle is the preferred database, but others such as open-source MySQL and Microsoft Access are sometimes used.
With researchers on both sides of the Atlantic, communication and knowledge management is critical. Morris' team has already created an e-collaboration toolbox with an e-collaboration program manager. But Novartis still largely depends on paper lab notebooks — for patent tracking reasons. Morris would like to build a comprehensive electronic laboratory environment in which "paper is just something you do for whimsical sake. Maybe that's too grandiose, but that's where we're thinking."
It also seems to be what Fishman is thinking. "Mark is a phenomenal technophile," Morris says. "He's asked general questions going across the life cycle about a compound, and then will want to drill all the way down, practically to the scientist who did it and what color hair he had that day. We want to be able to deliver the kind of interface [to do that]."
"One of the things that struck me was how readily Novartis wants to adopt IT in its infrastructure. It's a risky proposition to upgrade — there are all kinds of risks incorporating a new device or instrument in the environment. Novartis seems to be very willing and able to do that ... I've never been told, 'No, not yet, not this year.'"
The Novartis pipeline has rivals salivating with envy. From 2000 to 2002, the FDA approved 12 Novartis drugs, including Gleevec — a tally that equaled the combined approvals for GlaxoSmithKline, Pfizer, Merck, AstraZeneca, Aventis, Bristol-Myers Squibb, and Lilly. Last year, Novartis reported a net income of $3.8 billion on revenue of $24.9 billion.
"If you look at our pipeline, it's as good as anyone's in the industry," Kamb says. The company predicts it will have seven blockbusters by 2008 — perhaps more if persistent rumors of its interest in acquiring Roche pan out. (Novartis owns a 32.7-percent stake in the parent company of Hoffman-La Roche.) A merger would catapult Novartis into the champions' league of big pharmas, with sales in excess of $40 billion per year, behind Pfizer but ahead of GlaxoSmithKline.
Like all pharmas, Novartis faces intense scrutiny to justify the steep prices of new drugs. Vasella contends, "In one way, we are just businesses, no different from a Microsoft, General Electric, or PepsiCo. We sell products and want to make a profit. Without profits there would be no capital, no research, no investments, and eventually no new products."
Indeed, Novartis plowed $3.8 billion — 15 percent of sales — back into R&D last year. Kamb recognizes that, before long, he and his peers will need to produce "some successful drugs in clinical performance, but that is largely out of my control. We have a Phase III trial going right now — an anti-angiogenesis drug, an antagonist of the VEGF [receptor] — there's some Phase I data suggestive of efficacy, but we're not going to know until we unblind the study. That will have a big influence on the organization — it won't make or break, but it's a big bet with a big investment, big upside."
Talent and technology, science and surroundings bode well for Novartis, but ultimately, much depends on serendipity. Kamb isn't making any rash promises: "The jury's out as to whether we can come up with another Gleevec."
By Kevin Davies and John Russell
PHOTOS OF BUILDINGS FURNALD & GRAY; PHOTO OF KAMB BY FURNALD & GRAY; PHOTO OF ZEBRAFISH BY NATURE BIOTECHNOLOGY