Good Pill, Bad Pill: Science Makes It Hard to Decipher...

December 22, 2004

http://www.nytimes.com/2004/12/22/science/22pill.html

December 22, 2004

Good Pill, Bad Pill: Science Makes It Hard to Decipher

By GINA KOLATA

In one of the great examples of the mixed messages of science, the same

study that killed the blockbuster arthritis drug Vioxx after showing that it

had heart risks also found that the drug had a significant benefit: it

prevented precancerous colon polyps in some patients, one of the study's

principal researchers said.

But the drug's maker, Merck, and the researcher, Dr. Bresalier, said

that neither Merck nor the researchers had known that Vioxx prevented polyps

when Merck stopped the study and withdrew the drug from the market.

" At the time we made our decision to voluntarily withdraw Vioxx, the study

had not yet been completed, and efficacy results had not been disclosed to

Merck by the study's steering committee, " Loder, a Merck

spokesman, said.

Instead, because an independent board overseeing the study had reported that

Vioxx was associated with a twofold increase in heart attacks and strokes,

an unacceptable risk for otherwise healthy people, Merck announced on Sept.

30 that it was halting the 2,000-patient study. It withdrew Vioxx from the

market and stopped all other studies that asked if the drug could prevent

cancer. The company says it has no plans to bring the drug back.

Dr. Bresalier, who is a professor and chairman of the department of

gastrointestinal medicine and nutrition at M.D. Cancer Center,

explained that the polyp results emerged in data analyses only a few weeks

ago, showing that the study was " substantially positive. " He has not yet

presented the results to other scientists for review. The study was paid for

by Merck.

The company declined to comment on the polyp data except to say that they

would be presented at the American Gasteroenterological Association's

meeting in May and would be published.

But Dr. Bresalier said that while the final analyses were not in, it was

clear that markedly fewer patients developed precancerous polyps while

taking Vioxx and that no other study had shown such pronounced effects in

patients like those in this study, who previously had such polyps and were

at risk for developing more. Most colon cancer starts with polyps, although

most polyps do not become cancers.

Now, the mixed news about Vioxx raises questions of risk and benefit that

science simply cannot answer.

There is increasing evidence that not just Vioxx but similar drugs, like

Celebrex and Bextra, made by Pfizer, might increase the risk of heart

problems. But the risks of Vioxx and Celebrex came to light only because

they were tested in long-term studies to see if they could prevent cancer.

And if drugs like this do prevent cancer, which risk is worse, cancer if you

do not take them, or heart attacks if you do? Who decides?

In the meantime, scientists say they are reeling from the discouraging news

about drugs once thought to hold so much promise.

First Vioxx was withdrawn from the market. Then, last week, Celebrex turned

out to increase the risk of heart attacks in a similar colon polyp

prevention study. (Those polyp data have not yet been analyzed.) Bextra, it

turned out, increased heart attack risk in patients who had had heart

surgery. And on Monday, even Aleve, the over-the-counter pain reliever, fell

under suspicion. In a study asking whether it or Celebrex could prevent

Alzheimer's disease, Aleve patients had a 50 percent greater rate of heart

problems, which is a small increase that may not be statistically

significant. The study was halted anyway, in today's atmosphere of extreme

caution.

All this points to troubling realities in the calculus of risk versus

benefit.

What if you are at high risk of colon cancer? Do you take Celebrex? Or what

if you have arthritis and end up with a choice of a drug that can cause

ulcers and heart attacks or one that confers a greater risk of heart

attacks, but does not cause ulcers?

" This is where it all becomes murky, " said Dr. L. , the

director of the Cancer Research Center at Brigham Young University and a

discoverer of the COX-2 enzyme. The discovery led to the development of

Vioxx, Celebrex and Bextra, the drugs that block the enzyme.

And what if you are a drug company, wondering whether to pour money into a

long study of a drug that might prevent cancer? The Vioxx and Celebrex

stories illustrate the chance a company takes. Pfizer still sells Celebrex,

but stock prices in both companies plummeted with the news of their drugs'

risks, and Merck faces a raft of patients' lawsuits.

Yet cancer prevention trials take years and involve healthy people,

magnifying the risk that some drug dangers will emerge, scientists say.

" If you are talking about treating otherwise healthy people for years and

years and years with drugs, it is almost impossible to be sure that in some

relatively small fraction of patients, there won't be side effects, even

severe side effects, " said Dr. Bert Vogelstein, a colon cancer researcher

and investigator at s Hopkins University.

Now, Pfizer and Merck have seen what the consequences can be.

" I hope it doesn't prevent companies from trying to develop anti-cancer

drugs, " Dr. Vogelstein said. " But I fear it will. "

Celebrex is known to prevent polyps in patients with familial adenomatous

polyposis, a genetic condition that makes it inevitable that one day they

will get colon cancer. But even these patients are asking their doctors it

they should still take Celebrex. Or should they give it to their children,

who inherited the condition?

" I have patients right now who are struggling with this, " said Dr.

DuBois, a colon cancer specialist at Vanderbilt University School of

Medicine who has been a consultant to Merck and Pharmacia, a previous maker

of Celebrex. One family, Dr. DuBois said, learned just last week, when the

Celebrex news broke, that their child was affected.

Many scientists say that the decline and fall of the COX-2 inhibitors is

particularly poignant since these were drugs that once seemed to be the

perfect marriage of science and clinical medicine. They were designer drugs,

deliberately created to have the pain-relieving and anti-inflammatory

effects of aspirin without its side effects of ulcers and bleeding. And

laboratory research, animal studies and serendipitous observations of cancer

patients made researchers hopeful that the COX-2 inhibitors would be the

first safe and effective drugs that could actually prevent cancer.

The story began in the 1980's, when scientists studying cancer discovered

COX-2, which was activated when cells were inflamed or proliferating. It

turned out that aspirin and related drugs blocked COX-2, but they also

blocked a related enzyme, COX-1. And blocking COX-1 prevents platelets from

clumping and so can cause bleeding and stomach ulcers.

That led scientists, and drug companies, to what sounded like a brilliant

idea. Develop a drug that blocks only COX-2 and you could have aspirin's

pain-relieving and anti-inflammatory effects without its bleeding problems.

You might prevent cancer, by preventing cell proliferation. You might

prevent Alzheimer's disease, which has been associated with inflammation.

There were even hints that blocking COX-2 might prevent the bone

deterioration in osteoporosis.

" Gee, it looked great, " said Dr. Baron, a professor of medicine at

Dartmouth who has been a consultant to Merck.

And it looked as if at least some of the excitement would hold up. Vioxx and

Celebrex were developed, tested and shown to relieve pain and inflammation.

They became huge successes for Merck and Pfizer.

Then company scientists demonstrated that Celebrex could suppress polyp

formation in patients with familial adenomatous polyposis. In 1999, the Food

and Drug Administration approved Celebrex as a treatment for these high-risk

patients.

In the meantime, scientists were getting more and more evidence that drugs

that block COX-2 might also block cancers, including those of the colon,

breast, lung, bladder, skin and esophagus. Animal studies and

epidemiological studies looking at people who happened to be taking the

drugs for conditions like arthritis helped bolster the case - the drugs

appeared to protect against at least some forms of cancer.

Soon dozens of clinical studies of Vioxx and Celebrex got under way, with 20

Celebrex studies at the National Cancer Institute alone, involving people at

high risk for cancers of the lung, breast, skin, prostate, colon, mouth,

bladder and esophagus. The cancer institute also started 20 additional

Celebrex studies to see if the drug could help treat patients who already

had cancer.

Researchers were optimistic.

" It seems to be one of the best-documented cases that chemoprevention might

work for cancer, " Dr. Vogelstein said.

But all the while, some scientists had nagging concerns that COX-2

inhibitors might increase the risk of heart disease. Among them was Dr.

Garret A. FitzGerald, chairman of the department of pharmacology at the

University of Pennsylvania.

Dr. FitzGerald's line of questioning began in 1998, with a paper published

the week Celebrex was put on the market. In it, he showed that if he blocked

COX-2 in mice, blood vessels constricted and narrowed, and platelets, which

are involved in blood clotting, become active. " We saw this and said, 'This

is not good,' " Dr. FitzGerald said.

He continued to probe the biochemistry, but not everyone was convinced, and

even Dr. FitzGerald decided that the drugs' benefits - they relieved pain

and were less likely to cause stomach bleeding - outweighed for some

patients what he saw as their heart attack risks. But he was not surprised

when the colon cancer prevention studies showed that both Vioxx and Celebrex

were associated with an increased incidence of heart disease.

Some scientists point out, however, that no one has ever studied older drugs

that inhibit both COX-1 and COX-2 to see whether they also increase heart

attack risk. The Aleve data are preliminary, not conclusive, but they are

among the first to address the question for an older drug. Other such drugs

include Mobic, indomethacin, and diclofenac.

This raises the question, Dr. said, of whether all along people may

have been exposed to a slight heart attack risk in order to get relief from

the pain of arthritis.

" This may be something that arthritis patients have been facing perhaps for

decades and it hasn't been known, " he said.

And it may never be known, Dr. Bertagnolli, a cancer surgeon at

Harvard, said. " No one is going to take some poor arthritis patient and put

them on a placebo for three years, " to compare their heart attack rate to

others taking anti-inflammatory drugs.

For now, said Dr. Bresalier, science simply does not and can not have all

the answers. Every drug has risks and benefits, and often it is impossible

to know all of them even after a drug is being sold. But that does not mean

that drugs are bad or that federal regulators are lax.

" It's muddy, " Dr. Bresalier said. " Even people who are experts don't have

the answers. "

Corrections | RSS | Help | Back to Top

December 22, 2004