(FORTUNE Magazine) – As Bruce Roth moves about his small town of Plymouth, Mich., it's not unusual for people to buttonhole him in line at the supermarket, or corner him at a dinner party, and offer their gratitude. "I have neighbors coming up to me in church, thanking me for saving their lives," says Roth, vice president of chemistry at Pfizer's Ann Arbor laboratories. He's not just a local hero either. Last year some 44 million people around the world lowered their cholesterol by taking daily doses of the drug he invented 17 years ago, a chemical compound called atorvastatin calcium. In its six years on the market, the medicine--known by its more marketing-friendly moniker, Lipitor--has become the largest-selling pharmaceutical in history. Within the next few years it could very well become the world's first $10-billion-a-year drug.

If you've never heard of Lipitor or its connection to Pfizer, you're forgiven. With a wink and a nudge, Pfizer is linked in the national consciousness with its little blue impotence pill, Viagra. That drug received so much publicity after its release in 1998 that you'd think it would be the main force behind Pfizer's estimated $35 billion in 2002 revenues. But compared with Lipitor, the vaunted "Pfizer riser," with an estimated $1.75 billion in sales, is pretty puny. In 2002, Lipitor achieved estimated sales of $7.4 billion while commanding a 42% market share in this class of drugs, known as statins. Its closest rival, Merck's Zocor, held 32% of the market.

The outlook for statins couldn't be healthier. In the U.S. alone, some 52 million people need medical attention for high cholesterol, but only a third are getting it. As new patients seek out statins, Lipitor could well continue to gain share; Pfizer is a powerhouse marketer, and the drug will stay under patent protection until at least 2009. What's more, new markets could emerge: Lipitor and other statins are in clinical trials for possible use against Alzheimer's disease and multiple sclerosis.

Pfizer's status as the sultan of statins is even more remarkable when you consider that it was a latecomer to the cholesterol- reduction game. Lipitor entered the market a full decade after the first statins became available. The story of how Pfizer acquired the rights to an improved statin and turned it into the all-time biggest blockbuster is a tale of hyperaggressive marketing, deft timing, financial power, and plain dumb luck. It is also a story of missteps and missed opportunities by Merck, the company that pioneered statins.

The statin saga begins in an unlikely place--Japan. In 1971, Dr. Akira Endo, working for the small Tokyo drug company Sankyo, set out to create a cholesterol-lowering drug. Research a decade earlier had shown that most of the body's cholesterol is manufactured in the liver with the help of an enzyme known as HMG-CoA (the rest comes directly from food). Despite decades of research, no one had been able to locate a compound that inhibited the enzyme's production.

Endo was determined to do so. He and his team at Sankyo spent years sifting through more than 6,000 microbes, looking for one that naturally produced an HMG-CoA inhibitor as a defense against other microorganisms. Eventually he identified mevastatin, the world's first cholesterol-fighting compound.

Word of the breakthrough spread quickly in the global medical community--as did rumors that Endo's statin caused tumors, muscle deterioration, and sometimes death in laboratory dogs. The stories scared off researchers at several drug companies, but not P. Roy Vagelos, the chief scientist at Merck. His $1.5-billion-a-year company was betting heavily on applied research in molecular biology and biochemistry, and Endo's work intrigued him.

Starting in 1975, Vagelos made several trips to Japan; before long, Merck had successfully duplicated Endo's experiments. By 1978 its scientists had identified another cholesterol-lowering compound, lovastatin, and begun the long process of clinical trials needed for FDA approval. When lovastatin finally hit the market in 1987 under the brand name Mevacor, it was the first drug in its class. Vagelos had become CEO by then, and Merck had grown into one of the largest and most profitable drug companies.

Merck faced a dual challenge as the market pioneer. It needed to help educate the public about the dangers of high cholesterol in the blood. (Reducing heart attack risk in the mid-1980s was mainly a matter of avoiding saturated fat and of eating high-fiber foods like oat bran.) Merck also had to convince doctors that statins would extend their patients' lives. "Few doctors had experience with Mevacor, so they were reluctant to prescribe it," recalls James McKenney, a founder of the National Cholesterol Education Program at the National Institutes of Health. "For doctors, the questions were: How safe are these drugs, and can they change overall mortality?"

As Merck conducted research to prove that statins could stave off death, the public-awareness campaign took hold. Americans began drawing the connection between high cholesterol and heart attacks. Heart patients got to know their cholesterol numbers; people even learned the distinction between high-density lipoprotein (HDL), or "good" cholesterol, and low-density lipoprotein (LDL), or "bad" cholesterol. Food marketers responded with products like low-fat frozen pizzas and cholesterol-free potato chips. And rival drugmakers began to crowd into the statin market. Sankyo, for example, teamed with Bristol-Myers Squibb to market a drug called Pravachol.

More and more doctors were prescribing statins, yet questions lingered about the long-term benefits. Then, in April 1994, researchers announced the findings of the Merck-sponsored Scandinavian Simvastatin Survival Study, known as the Four-S. It had tested a new Merck statin on 4,444 patients with high cholesterol who had also experienced heart attacks. After five years of simvastatin use, the patients saw a 35% reduction in their cholesterol. Better yet, simvastatin had reduced the likelihood that a patient would die of a heart attack by 42%. "That was a golden moment for us," says Janet Skidmore, who led Merck's publicity campaign for the Four-S. Merck quickly won FDA approval for simvastatin, which it branded Zocor. With market-leading products and groundbreaking research, Merck soon made its name synonymous with cholesterol control. Zocor quickly rose to No. 1, and in 1995 both it and Mevacor were $1-billion-plus blockbusters.

Though clearly a godsend for Merck, the Four-S would also prove to be a curse. By lifting sales for the entire class of statins, it paved the way for Lipitor.

Back in 1982, when Mevacor was still five years from FDA approval, Bruce Roth was a 28-year-old postdoctoral fellow in the University of Rochester chemistry department. He had a fascination with statins and experimented with replicating their chemical structures. That spring he succeeded in synthesizing a statin that was very close to the one Sankyo's Dr. Endo had discovered a decade earlier. Roth's work wasn't groundbreaking, but his skill at piecing together and analyzing molecular structures attracted the attention of executives at Warner-Lambert's Ann Arbor labs. They invited him for an interview. "I thought it was intriguing," says Roth. "But honestly, I had never heard of the company."

He'd probably used its products. Once primarily a consumer goods company, Warner-Lambert was the maker of Schick razors, Visine, and Listerine. In 1970 the company had joined the ranks of Big Pharma by acquiring Parke-Davis. During Roth's visit to its labs, he was impressed by how Warner-Lambert was using cutting-edge computerized molecular modeling to speed the search for new drugs. He decided to sign on and by 1984 was head of an 18-scientist atherosclerosis discovery group.

Warner-Lambert had high expectations: It wanted to become a serious player in statins, and Roth's job was to make that happen. In August 1985 he synthesized Lipitor. At first, though, he didn't know what he had: Animal testing showed that while Lipitor worked, it wasn't any better than Merck's Mevacor. Roth knew Warner-Lambert wouldn't tolerate a me-too drug after all the resources it had poured into his project.

There were other concerns. Mevacor was already well along in its final round of clinical trials, but Roth didn't even have a process for manufacturing Lipitor in commercial quantities. "That took two years," says Roth, whose bosses grew increasingly anxious. By 1987, Warner-Lambert was floundering; its pipeline of new drugs had virtually run dry, and it had fallen from the fourth-largest drug company to the ninth. "People in '87, '88, and '89 were telling me it was too late for a cholesterol reducer," Roth says.

Whispers began to circulate that Warner-Lambert was close to terminating the project. It seemed clear that at best Lipitor would be the third statin in a still-unproven market. Roth argued for one last chance to test Lipitor in a human clinical trial. Ronald Cresswell, head of the labs, took the plea to Warner-Lambert's top managers. They agreed to fund one attempt to bring a cholesterol compound forward. If it failed, the company would end its statin effort. "We just rolled the dice," Roth says.

The dice came up sevens. A clinical trial of 24 Warner-Lambert employees showed that the earlier animal studies had greatly underestimated Lipitor. The drug's lowest dose for human subjects, ten milligrams, wound up reducing bad cholesterol by 38%, making it more effective than rival drugs at their highest FDA-recommended doses.

Suddenly Warner-Lambert knew it was on to something big. But to play catch-up in the market, it also knew it would need extra marketing muscle. Pfizer, then the No. 5 drugmaker, was known for its sales and marketing prowess and had no statin of its own. So when Warner-Lambert came knocking, Pfizer pounced at the chance. In 1996 the two companies entered an agreement to co-market Lipitor.

At first the partnership proved to be just what the doctor ordered. The marketers at Pfizer understood there might be advantages to Lipitor's late entry. By commissioning clinical trials, the companies could test their still-secret product against statins already on the market. The makers of those statins couldn't do the same because they couldn't get their hands on Lipitor.

In the spring of 1996 the results of a head-to-head trial landed on Pat Kelly's desk at Pfizer's New York City headquarters. The packet contained lots of paper, but one page stood out. It was a chart with color-coded curved lines, comparing Lipitor with each of the major cholesterol drugs--Merck's Zocor and Mevacor, Novartis's Lescol, and Bristol-Myers's Pravachol--in effectiveness at lowering cholesterol over time. "I will never forget seeing that chart," says Kelly, who was then a senior marketing executive and is now president of Pfizer's U.S. pharmaceuticals unit. Lipitor's curve began at a higher point than those of all the other medicines and showed a steeper upswing. That meant not only that Lipitor had a more effective starting dose, but also that it reduced cholesterol by proportions that the others couldn't. Pfizer marketers nicknamed the study the curves trial. "There it was in living color," recalls Kelly. "With that graphic, it was like, 'Aha! Now I know what this is about. We can communicate this!' " When the FDA approved Lipitor in January 1997, it allowed Pfizer and Warner-Lambert to include the curves data in the packet insert found in each prescription bottle of the drug. It was also part of the literature salespeople handed out to physicians.

In further preparation for the Lipitor market blitz, Pfizer conducted focus groups with doctors. The sessions uncovered a selling point that other statin makers had missed: Though doctors were prescribing statins in record numbers, they were still spooked at the idea of recommending high doses. The insight led Pfizer to push the FDA to approve a low starting dose for Lipitor. The agency agreed. It approved Lipitor for use in the range of ten to 80 milligrams a day. Other statins had been approved in the 20- to 40-milligram range. When doctors heard that, explains Antonio Gotto, dean of Cornell University's Weill Medical College, "they thought, 'If the FDA approved Lipitor up to 80 milligrams, ten milligrams must be really safe.' It was a major psychological advantage."

Pfizer's marketers now focused on toppling Merck's Zocor. To do so, they decided they needed an additional tactic: They would undercut Merck's price. "We were placing a high value on market analytics," says Mike Suesserman, Pfizer's director of marketing for Lipitor. "The pricing was based on feedback from doctors and managed care. We wanted to make sure Lipitor was within reach of all patients." (Today a month's supply of Lipitor costs about $66, vs. about $120 for Zocor.)

When Lipitor hit the market in spring 1997, it piggybacked on the work other companies had done to raise consumer awareness and doctors' confidence. By June 1998, Lipitor had garnered nearly 18% of the market, making it No. 2 behind Zocor, which held 37%. Wall Street analysts nicknamed Lipitor the turbostatin; between mid-1997 and mid-1998 the stocks of Pfizer and Warner-Lambert leaped 102% and 83%, respectively. The success of Pfizer's doctor-savvy marketing was especially frustrating for executives at Merck. "We were so busy patting ourselves on the back for doing the Four-S study that we didn't realize doctors were prescribing based on potency and price, not the outcomes that we showed in the study," one recalls.

By linking up with Pfizer, however, Warner-Lambert had gotten more than it bargained for. In late 1999, Warner-Lambert agreed to be acquired by American Home Products. Its hard-charging Lipitor partner wasn't about to swallow that. Breaking with genteel drug industry tradition, Pfizer made an unsolicited bid for Warner-Lambert, touching off a bidding war. After seven months, Pfizer won, paying $90 billion in the biggest drug takeover in history.

With Lipitor fully under its control, Pfizer went into overdrive in 2001 with direct-to-consumer advertising. Its campaign aimed at conveying two simple messages. The first: You don't have to be visibly unhealthy to have dangerously high cholesterol. The second: "Know your number"--that is, the level of bad cholesterol in your blood. A reading above 160 milligrams per deciliter of blood is considered risky, while a reading below 100 is optimal.

Pfizer's sales campaign in doctors' offices has been no less aggressive. Selling drugs to physicians isn't easy. To get their foot in the door, sales reps have to be almost as knowledgeable as doctors about health-care trends. So Pfizer continually trains and tests its 13,000 salespeople. A five-week boot camp for recruits includes courses in anatomy and physiology. Nothing is left to chance. The trainees go through weeks of simulated sales calls in a mock physician's office, built like a movie set on one of Pfizer's upstate New York campuses. On the simulation stage, former sales reps play harried and irritable doctors. Trainees are timed and judged on their ability to deliver a pitch for a Pfizer drug.

All this reflects the company's guiding sales principle: It wants doctors to think of Pfizer sales reps as vital suppliers not only of drugs but also of new, useful medical information. "What's important is that [the reps] get access to doctors who are incredibly crunched for time," says Richard Braaten, vice president in charge of training. "We want them to be able to give standup presentations in the middle of a hallway if they have to." The approach seems to work. In a national survey of doctors by industry consultant Verispan in 2002, Pfizer's sales force ranked as "most esteemed." That reputation helped make Pfizer's sales reps the most productive in the industry, with an average of 552 calls per year, vs. 409 for GlaxoSmithKline and 379 for Merck, according to Verispan.

The sales and marketing forces sometimes get carried away. Last October, Pfizer was forced to pay the U.S. government $49 million to settle charges that it overcharged Medicaid for Lipitor. Company officials say the incident occurred before the Warner-Lambert acquisition and happened on the Warner-Lambert side of the co-marketing deal. Also last fall the FDA asked Pfizer to stop running magazine ads that claimed inaccurately that Lipitor may not have the side effects of other statins. Pfizer changed the wording.

Even as Lipitor's sales continue to grow, Pfizer faces challenges. Concerns about statins' long-term effects haven't disappeared. In 2001, Bayer voluntarily withdrew its statin, Baycol, from the market after the FDA linked it to 31 deaths from muscle deterioration. And a study last year by Dr. Peter Langsjoen, a cardiologist at East Texas Medical Center in Tyler, Texas, found that two-thirds of patients showed signs of heart muscle weakening after only six months of statin therapy. The study is far from conclusive, but the FDA is reviewing Langsjoen's request for a government-funded study of long-term statin use. He also petitioned the FDA last May to order a warning on all statin pill bottles. Above all, people need to exercise caution, Langsjoen argues: "Doctors are prescribing statins with reckless abandon. They are really, really tricky drugs."

Pfizer's marketing team could also soon be put to the test. By the end of 2003, AstraZeneca plans to release Crestor, a statin that purports not only to lower bad cholesterol but also to raise good cholesterol far more effectively than other statins. Crestor has encountered snags and still needs FDA approval, but its lead marketer, Adele Gulfo, knows the competition well. She was part of the Pfizer-Warner-Lambert prelaunch marketing team for Lipitor. Because Crestor does things that Lipitor doesn't, says Gulfo, "this won't be a traditional Coke vs. Pepsi market war. This market is still in its infancy; 36 million Americans need cholesterol therapy, and we have a much more aggressive stance on how to treat it." For its part, Pfizer has no plans to change its marketing tack. Says Kelly: "Our approach will be the same--'Doctors, you know Lipitor. We've shown you the data. Why do you need anything else?' "

Even if Crestor turns out to be a hit, there's probably room for both drugs in this growing market. With baby-boomers aging, the global market is expected to expand from $18.8 billion in 2002 to $23.6 billion by 2007. That's not even counting the possibility that statins may prove effective against Alzheimer's disease and multiple sclerosis.

Despite the heady prospects, Lipitor won't power all of Pfizer's growth. Which brings us back to Bruce Roth. The inventor of the world's biggest drug has moved on from statins. He now manages teams in areas like psychotherapy and dermatology and is responsible for a sizable chunk of Pfizer's $5.2-billion-a-year R&D budget. Roth is no longer a researcher; he's a full-time manager who has been "studying failure," as he puts it. "Last year we made over 5,000 compounds. Only half a dozen of them will make it to clinical trials. I'm spending a lot of time trying to understand the failures so that we can increase our odds of success," he says. As to the high stakes and potential rewards of his work, he has a reminder every morning--when he takes his own daily dose of Lipitor.

FEEDBACK jsimons@fortunemail.com