(FORTUNE Magazine) – IT'S EXOTIC: a cancer drug made from the bark of a yew tree that grows only in a few areas of the Pacific Northwest and gives shelter to the rare northern spotted owl. It's one of the most expensive drugs ever: Three whole trees produce only enough for a single treatment; a pound of the stuff costs $250,000. It's not even clear how well it works: It has yet to save a single life. While it shrinks tumors in 50% of breast cancer patients and 30% of those with ovarian cancer, and even makes them disappear entirely in a handful of cases, the tumors come back just as soon as treatment stops. Several patients have died from its side effects. So why has taxol, the most ballyhooed drug since AZT, set off a breakneck race among pharmaceutical giants and biotech startups alike? The lure of a big payoff, of course. Even if taxol isn't quite a wonder drug, the National Cancer Institute (NCI) calls it the most promising anticancer compound to come along in 15 years. Industry analysts guess that Bristol-Myers Squibb, so far the clear leader in taxol development, could turn it into a $500 million to $1 billion product annually by the end of the decade. That would make it the biggest-selling cancer drug ever. Zola Horovitz, a Bristol-Myers vice president who runs the company's huge, top- priority project to mass-produce taxol, calls such numbers sheer speculation. But he adds that Bristol-Myers isn't worried about recovering its investment -- $100 million so far. ''Oodles of people are in, or would like to be in, the taxol game,'' says Neil A. Jans, vice president of Hauser Chemical Research in Boulder, Colorado, which has a contract with Bristol-Myers to collect yew bark and extract raw taxol from it. Bristol-Myers has largely locked up the domestic supply of trees through an agreement with the U.S. Forest Service and the Bureau of Land Management, and it has won ''orphan drug'' status for taxol as an ovarian cancer treatment from the Food and Drug Administration. That gives the company a seven-year monopoly on selling it for that purpose, once the drug gets FDA approval -- which is expected early next year. But plenty of other American, European, and Japanese companies are thought to be working in secrecy on taxol and related products. Says James McCamant, publisher of the Medical Technology Stock Letter: ''I've never seen so many people trying different approaches to make one thing.'' Rhone-Poulenc Rorer, a Franco-American drugmaker mostly owned by the chemical giant Rhone-Poulenc, claims to have solved the scarcity problem by deriving Taxotere, a close relative of taxol, from European yew needles and branches using a partly synthetic process. From the same amount of raw material, Rhone-Poulenc Rorer's method turns out larger quantities of what might prove to be a more effective drug than Bristol-Myers's. The technique also avoids chopping down yew trees, which U.S. environmentalists are trying to have declared a threatened species. A much smaller player, Escagenetics Corp. of San Carlos, California, is already making taxol in modest quantities with a method called plant cell tissue culture. Technicians take a piece of tissue from yew seedlings, grow it in a petri dish with various nutrients, and then use it to make taxol in fermentation tanks. CEO Raymond J. Moshy says he will be turning the stuff out in 100-pound batches by the end of next year. He claims his way costs one- tenth as much as extracting taxol from bark and doesn't require cutting down a single tree. Bristol-Myers will soon look at a sample of his taxol; the giant company is already supporting cell tissue culture work on the drug at tiny Phyton Catalytic Inc. in Ithaca, New York. Liposome Technology of Menlo Park, California, has a deal with Escagenetics to put taxol into tiny capsules that will steer it more precisely to cancer cells, thus reducing side effects. Other small companies including Polysciences Inc. of Warrington, Pennsylvania, and NaPro BioTherapeutics Inc. of Boulder are also extracting taxol from yew needles and branches, which are of course more readily renewable than entire trees. Even Bristol-Myers has no intention of relying on the existing stock of trees, which the Forest Service puts at 23 million, though that is really a guess. The company has signed up timber giant Weyerhaeuser Co. to plant millions of yew seedlings for future harvesting and is looking into making taxol from needles alone. Horovitz says Bristol-Myers should be able to do without yew bark within five years. Like Rhone-Poulenc Rorer, the company is banking on the partly synthetic production of taxol. Toward the end of this year, from a European supplier it declines to name, Bristol-Myers will be getting large amounts of a taxol chemical building block extracted from European and Asian yew needles and branches. At a New Brunswick, New Jersey, plant, it will then attach a chain of molecules to turn the stuff into taxol. WHAT ABOUT making taxol entirely synthetically, without using so much as a single yew needle? Stanford University chemistry professor Paul Wender and his ten-person research team, the academic front-runners in the field, expect to have wholly man-made taxol in hand by the end of this year. Wender cautions, though, that the feasibility of making taxol synthetically on a commercial scale remains unproved. While all the players find taxol's commercial possibilities irresistible, the longer-range reason the drug fascinates them is the unusual way it works -- which many scientists believe suggests that derivatives of taxol could work even better. Most cancer drugs keep cancer cells from reproducing by damaging their DNA. Taxol, by contrast, ''freezes'' those cells early in the process of division (see diagram). As a cell divides, a spindle of microscopic spaghetti- like tubes called microtubules emerges to help the process along. When the tubes contract, they act like tiny motors that pull the chromosomes apart so that half can go to each new cell. Taxol immobilizes these biological engines, paralyzing the tubes like so much stuck-together pasta. Unable to divide, the cell eventually dies. ''Taxol is a step forward -- but it's not a miracle drug and it's not a cure for cancer,'' says Horovitz. Given by itself, taxol, like most other cancer drugs, has severe side effects such as hair loss, lowered red and white blood cell counts, and a slowed heartbeat. Some patients, however, tolerate taxol better than existing chemotherapy agents. ''You talk to patients, especially after they have had other medications, and they say, 'Gee, I feel a lot better after taking taxol,' '' Horovitz adds. ''Even if they are not getting any response, they report feeling better. And the ones who are getting a response of course are ecstatic because their tumors shrink.'' Some patients suffer fewer severe gastric side effects than in conventional chemotherapy. That allows them to eat more or less normally -- a big plus. Taxol emerged in the late 1970s from a National Cancer Institute study of natural substances that might have cancer-fighting properties. Once Susan B. Horwitz and her associates at the Albert Einstein College of Medicine in New York City showed how taxol works, and the first reports of successful -- albeit temporary -- shrinking of ovarian tumors began to come in from Johns Hopkins Medical School, Bristol-Myers jumped in. The company, the No. 1 U.S. producer of cancer drugs (among its others: Paraplatin and VePesid), won a competition last year against Rhone-Poulenc Rorer and two smaller American companies, LyphoMed Inc. and Unimed Inc., for the right to enter into what's called a cooperative research and development agreement with NCI. Bristol-Myers's huge financial resources obviously helped. Bruce Chabner, director of cancer treatment at NCI, explained at the time: ''In our judgment Bristol-Myers Squibb was clearly the most qualified because of its extensive experience in the development of anticancer drugs and its ability to follow through to final new drug application status.'' The agreement made the company NCI's sole commercial partner for the clinical and commercial development of taxol. In return for Bristol-Myers's investment, NCI turns over to the company clinical data on the taxol trials it supports. Last year Bristol-Myers delivered to NCI nearly three pounds of taxol made from yew bark, which treated 600 cancer patients. The 40 pounds it expects to ship this year should go to some 8,000 patients. THERE'S NOTHING to stop another company from signing an agreement with NCI to supply a taxol derivative for treating ovarian or any other cancer or to develop taxol itself for other purposes. In fact, Rhone-Poulenc Rorer has just reached such an agreement for use of Taxotere with various forms of cancer. Although Rhone-Poulenc Rorer's clinical trials in the U.S. lag behind those sponsored by Bristol-Myers and NCI, Rhone-Poulenc Rorer will probably become Bristol-Myers's chief competitor in treating a number of cancers with taxol or taxol-like drugs. But not right away. Bristol-Myers has covered all the taxol bases -- from cornering the tree market to using all the natural and synthetic types of production. So for at least three or four years this company is likely to be the only big funnel dispersing taxol to cancer patients eager to try it.