A popular arthritis drug that seems to raise the risk of heart problems also seems to lower the risk of growths that lead to colon cancer, two large studies reveal.

The news on Celebrex poses a dilemma but also a potential prevention tool for people most likely to develop the deadly bowel disease — those with bad genes or many precancerous growths.

However, experts said the heart problems mean that the drug can't be recommended for preventing colon cancer in people at average or slightly higher risk.

"We are all concerned about cardiovascular events," said Dr. Nadir Arber of Tel Aviv Sourasky Medical Center in Israel, who led one of the studies.

The studies, presented Monday at a cancer conference in Washington, found that Celebrex cut the chances of developing precancerous growths called polyps by 33 percent to 45 percent in people who already had such growths removed.

They are the latest chapter in the saga of cox-2 inhibitors, painkillers that block a substance that causes inflammation and also is found in great supply in many tumors. Celebrex is the only such drug that remains on the market; Vioxx and Bextra were withdrawn over safety concerns.

In fact, a study testing Vioxx for cancer prevention led to the discovery of its heart risks in September 2004 and prompted researchers to suspend similar studies testing other cox-2 drugs against cancer.

Two of those suspended studies produced the news reported on Monday.

One, funded by the National Cancer Institute and Celebrex's maker, Pfizer Inc., involved 2,035 people in the United States, England, Australia and Canada. They were given either 200- or 400-milligram doses of Celebrex twice a day or dummy pills and then checked one year and three years later with colonoscopy exams.

More than 60 percent of those on dummy pills developed new polyps, but less than half of those on Celebrex did, said Dr. Monica Bertagnolli of Harvard-affiliated Brigham and Women's Hospital in Boston.

About 3.4 percent of those taking the drug had heart attacks, strokes or other serious heart-related problems versus 2.5 percent on placebo pills.

A second study, paid for by Pfizer, reached similar conclusions. Nearly 1,600 people in 32 countries who had previous polyps were given either 400 milligrams a day of Celebrex or dummy pills.

Three years later, new polyps had been found in 34 percent of those given the drug and 49 percent of the others.

Heart-related side effects occurred in 7.5 percent of those on Celebrex versus 4.6 percent of those on dummy pills.