Is Smithkline's Future In Its Genes?

R&D chief Poste is racing ahead in gene-based drugs

George Poste never lacked drive. The son of a British auto mechanic, he reached the Formula 5000 racing circuit by his early 20s. But as he was speeding around a German track in 1971, Poste had a horrendous crackup. In addition to broken bones, he suffered a crushed nerve in his larynx, forcing him to relearn to speak. It wasn't the 10 weeks in the hospital that ended his racing, though. "It was the realization that I had no chance of being a superstar," he says.

These days, Poste is shooting for a different kind of stardom. As research chief of London-based SmithKline Beecham PLC (SB), he closed a controversial $125 million deal in 1993 with Human Genome Sciences Inc. (HGS) in Rockville, Md., giving SB dibs on the largest database of human-genetic information in the world. Now, he's racing to capitalize on that. He wants to make genomics--the study of genes and their role in health and disease--the technology underlying SB's drugs, diagnostics, and consumer businesses. By doing so, he hopes to make SB an industry leader.

That won't be easy. The $11 billion Anglo-American company is only the world's 10th largest in terms of drug revenues, down from No.3 in 1989, after a wave of industry mergers. Although SB has 4,200 scientists and R&D staffers divided between Britain and the U.S., its $1 billion R&D budget is shrinking compared with the outlays of its rivals. Some insiders even wonder whether the outfit is big enough to go it alone.

That's why Poste, a confessed caffeine addict who drinks up to 30 cups of coffee a day, is pushing SB's scientific pedal to the metal. He figures SB has two to three years left to wring value from HGS's genetic data before rivals catch up. So he's forging alliances and shifting his core research and development group into high gear. Even before the genomics kicks in, SB is producing at least two novel drugs and one vaccine per year. It has 23 compounds in late stages of human testing or awaiting approval--more than any other company. That has pushed SB's stock up 50% since last March--to $10.92.

JUMPING CURBS. For Poste, that's not enough. "Is this a pressure-cooker organization? Unequivocally. But that's the landscape of the times," says the research chief, who has been known to fly into eye-bulging rages and terrify passengers when he's behind the wheel by jumping curbs to get to his destination faster.

Poste, 51, has been as quick to shift gears in his career as in his car. He earned a degree in veterinary medicine in 1966. But after spending eight weeks tending to cows and sheep on the English-Welsh border, he called it quits. He got a PhD in virology at the University of Bristol in 1969, then taught at the State University of New York at Buffalo. In 1980, SmithKline and French Laboratories lured him away. Nine years later, its parent SmithKline merged with Britain's Beecham, and Poste became No.2 in the combined lab, rising to become research chief in 1991.

A voracious reader with a near-photographic memory, Poste also has persuasive powers that even competitors salute. "He's just a phenomenal speaker with a real vision of what the world is going to look like," says Randall W. Scott, chief scientific officer at Incyte Pharmaceuticals Inc. in Palo Alto, Calif.

Poste's forcefulness served him well in the early 1980s when he persuaded his bosses to let him recruit the National Institutes of Health's Martin Rosenberg, a protein expert. The result in 1989 was Engerix-B--one of the first genetically engineered hepatitis-B vaccines, with $613 million in sales last year.

Poste took another flier on Human Genome Sciences. In mid-1993, he got SB's board to pay $125 million for 7% of the company and exclusive commercial access to most of its data. Rivals were skeptical. At the time, the data's value was unclear. Glaxo, for one, declined to invest in HGS, says Colin Dykes, its genomics chief in Britain, because of the daunting "struggle to make sense out of mountains of sequenced information."

Poste coped with the data overload by hiring "bio-informaticians," who are equipped with powerful computers to help locate genes by searching for patterns in DNA sequences. With 30 such scientists already, SB aims for 50 by yearend. It has just nabbed one of the best--Randall F. Smith of the Human Genome Center at the Baylor College of Medicine.

Genomics has yet to lead to a single blockbuster drug. Yet Poste has based 25% of SB's drug-discovery programs on it and expects to base all of them on it by the year 2000. The idea is that systematic database research can lead to the discovery of genes with previously unknown roles in disease. Those discoveries can then be used to develop drugs to act on the genes or their corresponding proteins. It's a gamble. Says Ian Smith, a former Beecham scientist who's now a Lehman Brothers Inc. drug analyst: "If he's wrong, the company could run into a lot of dead ends."

For now, SB has a clear lead. HGS has sequenced about 850,000 fragments of DNA, which represent pieces of about 85% of the genes in the human body. In contrast, a Merck-funded group at Washington University in St. Louis had about 355,000 sequences through January. SB claims to have a two-year lead, but Merck boasts of wider collaborations in the academic community.

Putting its genomics expertise to work, SB is racing ahead with a drug developed from analysis of a gene that causes a form of osteoporosis (diagram). The compound may enter clinical trials within a year, although a commercial drug is still years away, says SB's Rosenberg. By the turn of the decade, SB hopes to market tests that would diagnose diseases by detecting changes in the way genes are turned on and off. While diagnostics itself is a small market, the underlying science could also lead to new treatments.

Meanwhile, SB and its collaborators are patenting genes like crazy--filing nearly 200 of the 450 human-gene-based patent applications that have been filed so far in the U.S. One is for a gene involved in atherosclerosis. HGS's database compressed what would have been a yearlong gene hunt by 5 to 10 researchers into a 24-hour computer exercise.

At the same time as SB hunts for genes, it is developing a fast way to test the effectiveness of both gene-based and conventional drugs. That work is being done by Orchid Biocomputer, a joint venture set up last fall by SB and the David Sarnoff Research Center in Princeton, N.J. By 1998, Orchid wants to put 10,000 microscopic test tubes, along with computer-controlled pumps and sensors, on chips the size of business cards. In a few hours, the chips would make 10,000 chemicals and then screen them against biological targets. Poste figures SB's chemists will be able to test 5 million chemicals a year, 500 times as many as now.

"TOUGH BOSS." Poste wants to shorten the time it takes SB to bring drugs to market. Every week of development costs the company $11 million in lost revenue on a big-selling drug. His goal: to squeeze development time to just under 5.5 years--down from 7.8 now and more than 9 in 1991. "He's a very tough boss who sets tough deadlines," says Paul Nicholson, SB's senior vice-president for worldwide development.

That's for sure. "People know when they give me a stupid or superficial answer," he says. Since becoming boss, he has cut the R&D organization by nearly a quarter, axing 1,300 people who he felt were redundant or had not signed on to his vision. "This hasn't always made me the most popular guy in the organization," he concedes.

Still, he has invigorated the company. Before he took over, "they had all that biology but no breakthroughs," says biotech pioneer George B. Rathmann, founder of Amgen and now chairman of ICOS Corp. in Bothell, Wash.

Five years from now, Poste sees himself spending more time at his house in the Arizona desert, writing about technology or driving the Humvee that he recently mired in three feet of water in Arizona's Verde River. With his drive, five more years may be all the time Poste needs to make SmithKline Beecham a genomic star.

Before it's here, it's on the Bloomberg Terminal.