More than 70 years ago, a farmer approached scientists at the University of Wisconsin asking them to solve a mystery involving spoiled sweet clover, a dead heifer and bovine blood that wouldn't clot.

What the researchers found -- that the rotting clover triggered a chemical reaction that caused internal bleeding -- eventually led to the creation of warfarin, an anticoagulant that has played a key role in preventing debilitating and fatal blood clots.

Over the years, warfarin has been the target of companies that believed they could make a better, more convenient drug, and capitalize on a market where many patients must take anticoagulants for the remainder of their lives. Warfarin pills are sold generically and by Bristol-Myers Squibb ( BMY) under the brand name Coumadin.

"Anything that could supplant Coumadin would be widely used," says Albert Rauch, an analyst with A.G. Edwards. "They could expand the market."

Several companies hope to do just that. Currently, Eli Lilly ( LLY) Germany's Boehringer Ingleheim and Bayer ( BAY), and the Anglo-Swedish firm AstraZeneca ( AZN) have drugs in clinical testing.

However, many challengers have failed, and warfarin has proven durable.

The key issue is "to design something that can be taken by mouth and efficiently absorbed into the blood stream as a therapeutically effective agent," says Dr. Thomas Ortel, director of the hemostasis and thrombosis center at the Duke University School of Medicine. "New drugs have to be looked at carefully because patients may be on them for a long time, and long-term side effects may not be identified with relatively short clinical trials."

The most notable recent failure was Exanta from AstraZeneca. In clinical trials, the Exanta pill was as effective as warfarin in preventing clots and strokes. Exanta was easier to administer, and it interacted with far fewer drugs.

Unfortunately, clinical trials also revealed that Exanta raised liver enzymes, a signal of possible damage to the organ. The liver results were so strong that an FDA advisory committee opposed the drug in September 2004. A month later, the FDA rejected it.

Although the European Union had approved for Exanta for limited uses -- short-term clot prevention in patients having elective knee- or hip-replacement surgeries -- AstraZeneca recently decided to pull the drug , saying updated research showed that the risk of liver damage was too great.

Effective, Yet Vulnerable

Warfarin is an inviting target for competitors because patients must have their blood tested regularly to monitor the drug's level. If the warfarin level is too low, patients risk developing clots. If it's too high, they can have internal bleeding. A different form of warfarin has been sold as rat poison for nearly 60 years.

Acceptable levels are tough to maintain because warfarin reacts with many drugs and foods. The Coumadin label lists 60 classes of drugs and 120 individual drugs that can exacerbate the anticoagulant's effect. The label also identifies 28 classes of drugs and 43 drugs that can suppress Coumadin's action. Twenty-two drugs can either increase or decrease the Coumadin effect.

Many herbal medications also interact with it, and patients are counseled to beware of certain foods, like broccoli, Brussels sprouts and spinach. These foods are high in vitamin K, which the body uses to make clotting factors. Warfarin causes a vitamin K deficiency, reducing the body's blood-clotting ability.

Several injectable anticoagulants are on the market, but warfarin's big selling point is that it's a pill. Companies have had trouble creating clot-preventing pills because they haven't figured how their compounds can be adequately absorbed into the bloodstream.

Warfarin also endures because it delivers a special type of clot prevention. There isn't any one-size-fits-all anticoagulant -- a drug that works well in arteries won't be as effective in the veins, where the blood flows more slowly.

For example, the pill Plavix, from Sanofi-Aventis ( SNY) and Bristol Myers-Squibb, prevents clots in arteries by discouraging blood platelets from sticking together. (Aspirin has the same effect.) Platelets are specialized cells involved in normal clotting. The danger occurs when the platelets coagulate and latch on to a protein called fibrin, Rauch says. In the veins, dangerous clots are caused primarily by fibrin.

Warfarin works best in the veins, which is why it's prescribed to prevent deep vein thrombosis. A.G. Edwards' Rauch says there are more than 2.5 million cases of these lower-leg clots each year in the U.S., causing more than 50,000 deaths.

Additionally, Warfarin is the preferred clot-preventer for people with atrial fibrillation, an erratic heartbeat that impairs normal pumping by the heart's upper chambers. Patients with the condition risk having blood coagulate in the chambers that could eventually be sent as clots to vital organs. Warfarin also is the blood thinner of choice for people with artificial heart valves.

"People have looked for years to see if aspirin would work as well as warfarin in certain conditions," says Dr. Ortel. "In many cases, for example atrial fibrillation, it isn't as effective as therapeutic warfarin."

Potential Competitors

Given the history of unsuccessful challenges, it's hard to handicap the race for a successor to warfarin, which gets its name from its patent holder. Even under the best-case scenario, a competing drug probably wouldn't reach the market until 2009.

AstraZeneca is working on AZD0837, for which early clinical trials suggest fewer side effects than Exanta. Like Exanta, AZD0837 inhibits thrombin, an enzyme involved in the clotting process. However, AstraZeneca reported in January that the drug doesn't remain in the body long enough to become a once-a-day product. The company said it will study making an extended-release version, but the extra research will push back late-stage clinical trials by two years.

Rauch says the Boehringer Ingleheim drug BIBR-1048 appears to be in the lead, only because it's in phase III trials, the last stage of clinical testing before companies seek regulatory approval. The drawback is low bioavailability, a measurement of how well the body absorbs a drug.

A drug with low biovailability requires administering more of the drug, and it also leads to more variability in how the drug is absorbed.

Rauch says midstage clinical data from Eli Lilly look "pretty impressive" for its drug LY517717, to prevent clots in patients who have undergone orthopedic surgery.

Bayer, which is working with Johnson & Johnson ( JNJ), has completed midstage clinical trials with BAY 59-7939 to prevent clots in veins after orthopedic surgery. The Bayer data look "good," says Rauch, who has a buy rating on Lilly but doesn't cover the other companies mentioned in this article. His firm doesn't have an investment-banking relationship with it.

Bayer has started a phase III test of the drug to prevent clots in patients undergoing hip or knee surgery, and it plans a late-stage trial, in a few months, for atrial fibrillation patients. The Bayer and Lilly compounds inhibit Factor Xa, an enzyme involved in blood clotting.

Another developer of a clot-preventing pill is Emisphere Technologies ( EMIS) of Tarrytown, N.Y. Emisphere has completed phase II tests on a pill version of heparin, a common injectable anticoagulant.