News FocusClinical Trials

Nail-Biting Time for Trials of COX-2 Drugs

See allHide authors and affiliations

Science  03 Dec 2004:
Vol. 306, Issue 5702, pp. 1673-1675
DOI: 10.1126/science.306.5702.1673

Preliminary studies suggest that the COX-2 inhibitor Celebrex may stem cancer and Alzheimer's disease, but testing these possibilities has just gotten tougher

Psychiatrist John Breitner was in a hotel room in Sun City, Arizona, when he heard the news on television. It was 30 September, and CNBC was reporting that the COX-2 inhibitor Vioxx would be yanked off the market by its maker, Merck, after experts saw a frightening increase in cardiovascular side effects. Breitner's own heart skipped a beat. In an instant, he realized that his effort to stop Alzheimer's disease using Celebrex, a Vioxx competitor, had just gotten trickier.

Similar but different.

The withdrawal of Vioxx has put Celebrex in the spotlight.

CREDIT: MARY ALTAFFER/ASSOCIATED PRESS Breitner, an expert on aging based at the University of Washington, Seattle, is one of dozens of researchers exploring whether COX-2 inhibitors can do more than they were designed to do—ease the painful inflammation of arthritis. Over the years, animal studies have suggested that these medications, along with more traditional nonsteroidal anti-inflammatory drugs (NSAIDs), may be able to lower the risk of cancer and reduce inflammation suspected in Alzheimer's.

In the past few years, scientists have launched one study after another to put these hopeful ideas to the test. The pace picked up after the U.S. Food and Drug Administration (FDA) confirmed data in 1999 showing that Celebrex reduces intestinal polyps in patients with familial adenomatous polyposis, a hereditary condition that leads to colon cancer. Excitement has focused on COX-2 inhibitors because they are believed to be less likely than NSAIDs to cause stomach problems, a big drawback in long-term prevention trials.

Few Vioxx prevention studies have been conducted or were planned, researchers say, partly because Merck was less willing than Pfizer, the maker of Celebrex, to donate a COX-2 inhibitor to such trials. The result is that about 10,000 volunteers are participating in or being recruited for Celebrex studies, but scientists can't say for certain whether the drug shares Vioxx's hazards. Now trial managers are debating the risks, reassuring study participants, and keeping a hand on the emergency brake just in case.

The last 2 months have been nail-biters for these researchers and their funders. The concern is heightened because in most of these trials, volunteers are healthy, and although many are at risk, not all will develop disease. “This is not fun for anybody,” says Curtis Meinert, chair of the steering committee for the Alzheimer's Celebrex trial and a clinical trials expert at Johns Hopkins University in Baltimore.

“Obviously, I was concerned” about the Vioxx announcement, says Jenny Mao, a pulmonologist at the University of California, Los Angeles, who's enrolling 180 former smokers in a lung cancer prevention trial that's testing Celebrex. Her biggest fear, she says, is that Celebrex, too, will be pulled, and that “all this work over all these years would go down the drain.”

Like most researchers, Mao believes that Celebrex doesn't induce the heart attacks and strokes seen with Vioxx. Although both drugs are COX-2 inhibitors, Vioxx is a more potent blocker of the COX-2 enzyme—a potential source of its problems—and also has a longer half-life. In addition, some cardiologists had warned for years of Vioxx's cardiovascular risks.

But a heightened level of scrutiny brought down Vioxx: The trial that persuaded Merck to withdraw its drug lasted 18 months and included 2600 people—longer than any single large, published Celebrex study. The purpose was to test whether Vioxx could prevent colon polyps—a precursor of cancer—in those at risk for developing them. Among volunteers on Vioxx, 3.5% suffered heart attacks or strokes, compared to 1.9% on placebo.

So far, “the data that are available … don't show the same” risks for Celebrex, says Meinert. “That,” he hastens to add, “isn't proof they don't exist.”

If there are cardiac problems, they might be hard to detect; researchers are straining to catch warning signs. The largest and longest running Celebrex prevention trial, a 2000-person study looking at the reappearance of colon polyps in patients with a history of them, has added a cardiovascular expert to its data safety and monitoring board (DSMB). In a meeting after Vioxx was withdrawn, DSMB members pored over trial data and agreed that the trial should continue, says Ernest Hawk, a chemoprevention expert at the U.S. National Cancer Institute who is program manager for the NCI-funded trial. NCI has also created a “cardiovascular adjudication process,” essentially asking a group of cardiologists to review and classify every potential cardiovascular event. Although DSMBs overseeing Celebrex prevention trials have been on heightened alert, and many have added a cardiac expert to their ranks, they “have not halted the trials or demanded changes to them based on cardiovascular risk,” says Hawk. (A Pfizer spokesperson confirmed that the company is not running any prevention trials with Bextra, its other COX-2 inhibitor.)

Alzheimer's prevention trials face challenges, too. Breitner's 2500-person study, the Alzheimer's Disease Anti-Inflammatory Prevention Trial (ADAPT), funded by the U.S. National Institute on Aging (NIA), uses Celebrex. All subjects must be at least 70 years old, putting them at a high risk of heart disease to begin with. Susan Molchan, NIA's program director for Alzheimer's disease clinical trials, contacted FDA after Vioxx was pulled off the market “to see if they had information” about Celebrex “that they could share,” she says. “They confirmed we weren't missing any information.”

NCI is trying to improve monitoring of Celebrex studies, according to Hawk. Meinert has urged agency officials to meld safety data from all the trials, making small signals easier to detect. “Trials are fairly weak instruments at finding adverse events,” he says, especially if they occur infrequently. “We need to have a better model, in my opinion, with regard to the harvest of safety data” among different trials studying the same drug. NCI's Hawk confirmed that the institute is speaking to Celebrex investigators about better ways to evaluate cardiac safety.

As scientists probe the Vioxx-Celebrex relation, they find that study participants are often primed with questions. “I've conducted town hall meetings for patients,” says Peter Lance, a gastroenterologist at the University of Arizona in Tucson, who's heading a 1600-person colorectal adenoma prevention trial involving Celebrex and the mineral selenium. Several dozen attended recent meetings in Tucson and Phoenix, where Lance explained that, thus far, there has not been an “imbalance” in cardiac problems among those taking Celebrex. “We're taking otherwise healthy people and asking them to take a medication or an intervention for which we don't have scientific evidence” of a clinical benefit, he says. “Our thoughts about safety are very intense.”

In the ADAPT trial, says Breitner, “we have people who are being advised to drop out by their physicians” and patients “who say they were going to enroll but aren't. We're definitely taking a hit from this.” Between 20 and 50 participants have refused to continue taking study medications (Celebrex, naproxen, or placebo), Breitner adds. To keep enrollment steady, ADAPT's coordinators have sent information about Celebrex, in lay language, to field sites. Although Breitner agrees that more information about Celebrex's long-term cardiac effects are needed, he doesn't think it poses anything like the risk of Vioxx: “I don't think that I'm running a trial where we're poisoning people.”

Many other trials haven't suffered much. UCLA's Mao says her staff was far more concerned with how the Vioxx withdrawal might influence their study than were participants. Other trials, moreover, include patients with such a high chance of cancer that cardiac risks pale in significance. For example, the 360 patients to be enrolled in the oral cancer prevention study headed by Scott Lippman of M. D. Anderson Cancer Center in Houston, Texas, will have a 60% chance of developing cancer in the next 3 years.

Celebrex researchers are hopeful that regardless of whether these trials show any effect on cancer or Alzheimer's risks, they will answer once and for all the question that's lingered since boxes of Vioxx were shipped back to Merck: whether Celebrex shares Vioxx's downside, and to what degree. Says Mao: “We'll keep our fingers crossed.”

View Abstract

Stay Connected to Science

Navigate This Article