Johnson & Johnson (JNJ) is placing bets on a new drug-releasing stent it hopes will vault the company back atop the $4 billion market for the tiny heart devices while it cancels plans for a different stent with more muted prospects.

The company is a stent-market pioneer but has seen its position drop sharply thanks to the rise of newer devices, especially a stent made by Abbott Laboratories (ABT) called Xience. In a show of support for a still-unproven device called Nevo, J&J's Cordis stent subsidiary will take on Xience in a big new study.

The randomized, 2,000-patient international study will pit Nevo, which is still being researched and isn't on sale anywhere, against what has become the top-selling stent in both Europe and the U.S. Cordis is replacing a just-started, non-randomized study on Nevo while saying it has a strong enough impression thus far to challenge a popular product.

The new study will help support an eventual U.S. regulatory filing. The beginning and end dates aren't yet set, but Cordis will know more after it meets with regulatory agencies to finalize the study design.

The move has risks, as Abbott's stent demonstrated its mettle by besting a blockbuster device from Boston Scientific Corp. (BSX) in a key study.

Campbell Rogers, the chief scientific officer at Cordis, called Xience a "formidable" product. But "we have a lot of confidence in Nevo," he added.

J&J doesn't yet have results from a smaller Nevo study that could lead to European approval, but has talked to doctors using the stent and has solicited input from leading cardiologists.

Shares of Dow Jones Industrial Average component J&J recently traded up 64 cents, or 1.3%, to $51.28 on a day of broader market gains.

Abbott spokesman Jonathon Hamilton said J&J's move isn't surprising, given Xience's leadership position, and that Abbott expects more head-to-head matchups.

While Cordis boosts Nevo's profile, it has also decided to stop developing Cypher Elite, which is a next-generation version of the Cypher stent J&J has sold domestically for six years. It is less likely to make a splash in an increasingly crowded U.S. market where Boston Scientific, Abbott and Medtronic Inc. (MDT) have all released new stents since early 2008.

Nevo isn't designed as an incremental update. The coated stents available today have medication attached by polymer that releases to keep propped-open heart arteries from closing off again. Nevo's drug is loaded into tiny reservoirs rather than coated on the surface, and both the drug and polymer are designed to be totally gone within three months. It's thought that the drug is only needed initially, and polymers can be irritants.

This technology may sound familiar to people who remember Conor MedSystems, a small stent developer J&J purchased in early 2007 for $1.4 billion. Conor launched a stent with this technology in Europe, but J&J pulled it from the market soon after buying Conor when an important trial failed.

That sapped J&J of its near-term ability to ward off rising competition, and the effects have shown. In 2008 J&J's coated stent sales fell 23% globally and nearly 39% in the U.S., the world's biggest market.

J&J has long maintained that it was more interested in Conor's technology than the failed device, and Rogers stressed that Nevo is a totally redesigned product.

It appeared that not enough drug, the same one used on Boston Scientific's Taxus stents, was used on the failed Conor stent. But the overhauled Nevo stent uses the same drug as J&J's Cypher devices, among other changes.

Abbott's Hamilton noted the company is working on a non-metal stent that will be completely absorbed by the body, rather than just shedding its drug and polymer.

J&J hopes to file for European Nevo approval by the end of this year, suggesting potential approval in 2010, if a study pitting Nevo against Boston Scientific's Taxus looks good. Six-month data will be presented at a European heart conference in May.

The company then plans to file for U.S. approval in late 2011, indicating it will battle domestically with the market's oldest coated stent - the original Cypher - for several more years. A non-randomized study of about 1,000 patients will serve as the "pivotal" trial aimed at supporting the eventual U.S. application.

Another U.S. study of about 2,000 patients, designed to compare clinical outcomes in Cypher patients who used anti-clotting drugs for different durations, will serve as the control arm for the Nevo stent's pivotal U.S. study. Cordis is working to expand Cypher's approved uses in the U.S. and is also working on versions designed for small and large arteries.

-By Jon Kamp, Dow Jones Newswires; 617-654-6728; jon.kamp@dowjones.com