Pfizer

(PFE) - Get Report

says two studies show the arthritis drug Celebrex helps halt the growth of precancerous colon polyps that can lead to cancer. But tests also confirm that the drug increases the risk of cardiovascular problems, again illustrating the dilemma of balancing risks vs. benefits.

The company continues to review the test data and will send final study reports to the Food and Drug Administration and foreign regulators.

"We haven't made a decision one way or the other" about seeking FDA approval for additional uses, Gail Cawkwell, Pfizer's senior medical director for Celebrex, said in a telephone interview with

TheStreet.com

.

Pfizer said Monday that the risk for cardiovascular damage detected in the two studies "is consistent with the current warnings for cardiovascular risk" on the Celebrex label.

Celebrex now carries a black box warning -- the toughest alert required by the FDA -- pointing out that Celebrex may carry an increased risk of blood clots, heart attack, stroke or death. The warning applies not only to Celebrex, which is called a COX-2 inhibitor, but also to older pain relievers called nonsteroidal anti-inflammatory drugs, or NSAIDs.

The black box warning came as the result of FDA advisory committee hearings and a review following

Merck's

(MRK) - Get Report

decision in September 2004 to remove its COX-2 pain reliever Vioxx from the market.

Merck acted because a clinical trial revealed

a higher cardiovascular risk among patients taking the drug for more than 18 months. Ironically, Merck's decision was based on a clinical trial testing if Vioxx could reduce colon polyps.

Last April, Pfizer acceded to an FDA request

to suspend sales of another COX-2 drug, Bextra. The FDA said the drug's risks outweighed its benefits. Pfizer strengthened its Celebrex label in July.

The colon-polyp test results were released at a meeting of the American Association for Cancer Research in Washington. One of the studies was cosponsored with the National Cancer Institute and the other was sponsored solely by Pfizer.

In 2004, preliminary results for the cosponsored study revealed "a statistically significant increase" in cardiovascular risk among people taking Celebrex compared with those getting a placebo, Pfizer said. The other test didn't show a statistically significant difference.

"A new broader analysis of serious and other cardiovascular events

for both studies found more cardiovascular events with Celebrex compared to placebo," Pfizer said Monday. "These results are consistent with the current warnings on cardiovascular risk with the Celebrex label."

Pfizer also is undertaking a Celebrex cardiovascular risk study of 20,000 patients worldwide, led by the Cleveland Clinic. The study will take 18 months, and patients will be followed for two years. It will track heart attacks, strokes and deaths among patients taking Celebrex and two older, now-generic drugs, naproxen and ibuprofen. Naproxen is best known as prescription-strength Naprosyn or over-the-counter Aleve. Ibuprofen is best known as Motrin.

The controversy over COX-2 drugs caused a big drop in Celebrex sales to $1.73 billion last year vs. $3.3 billion in 2004. Bextra had $1.29 billion in sales in 2004, but the cost of suspending the drug led to a loss of $61 million last year.

Celebrex has been approved by the FDA for many uses, the most notable being osteoarthritis, rheumatoid arthritis, a form of arthritis called ankylosing spondylitis, acute pain and menstrual cramps.

The drug also is approved for patients with familiar adenomatous polyposis, or FAP, a rare genetic disease that causes the rapid development of colon polyps. FAP patients have a high risk of colon cancer.

The latest studies looked at how Celebrex affects the development of sporadic precancerous tumors in the colon. Both tests were scheduled to run 36 months, but patients stopped receiving Celebrex after 33 months when the high rate of cardiovascular risk was detected in the National Cancer Institute-Pfizer test.

The risk was two to three times higher among recipients of Celebrex compared with those taking the placebo, depending on the Celebrex dose. The study tested 2,035 patients who received a placebo, 400 milligrams of Celebrex daily or 800 milligrams of the drug each day.

Sixty-one percent of the placebo patients developed one or more benign colon tumors called adenomas. For patients taking Celebrex, the rate was 42% for 400 milligrams and 37% for 800 milligrams. "Overall, this represents up to a 45% reduction in the development of precancerous polyps," Pfizer said.

Pfizer tested 1,561 patients in 32 countries in the other test, comparing those given the placebo with those receiving 400 milligrams of Celebrex. Among placebo patients, 49% developed precancerous polyps while 34% of the Celebrex patients developed the polyps. The Celebrex recipients had a 20% higher cardiovascular risk, but the rate wasn't statistically significant.

The normal Celebrex dose for osteoarthritis is 200 milligrams a day. The label says 400 milligrams daily can be applied to some of the other indications. For FAP patients, the recommended dose is 400 milligrams twice a day.

The two test results released Monday also provided what Pfizer's Cawkwell called "not definite but encouraging" news about Celebrex and gastrointestinal problems. She said the results will be part of Pfizer's future discussions with the FDA.

In both trials, "there were no differences" in the rates of gastrointestinal side effects when Celebrex was compared with the placebo, a Pfizer press release said. "However, the trials were not designed to examine gastrointestinal safety, and there were too few safety events to draw definitive conclusions."

Celebrex's label has a black box warning for gastrointestinal problems, such as bleeding and ulcers, especially among the elderly. Older pain relievers have the same warning. For many years, developers of COX-2 drugs tried to convince the FDA that their products were safer on the stomach than were the older medications, but only Vioxx's label got approval to say the drug caused fewer side effects than naproxen.