• 4/25/2005
  • Bethesda, MD
  • Cheryl A. Thompson
  • Am. J. Health Syst. Pharm. 2005; 62(9): p. 890-894

While September 30, 2004, may be remembered as the day that clinicians and patients started reconsidering the available pain relievers, the negative news about rofecoxib sent a shock wave through the world of preventive oncology. December 17 brought bad news about celecoxib.
Researchers had been exploring the potential of long-term therapy with rofecoxib or celecoxib, both cyclooxygenase-2 (COX-2) inhibitors, to reduce the risk of colorectal cancer.

What the researchers found first was that the drugs apparently increased users’ risk for myocardial infarction, ischemic stroke, and other cardiovascular events.

The ability of COX-2 inhibitors to obstruct the activity of the inflammation-promoting enzyme and also interfere with inflammatory mediators such as interleukin-1 and nuclear factor kappa B had made that type of drug a potential chemopreventive for cancer, according to Ernest T. Hawk of the National Cancer Institute (NCI).

Buttressing this line of research was evidence that an excess of COX-2 played a role in the invasive and earlier stages of cancer at a large number of sites in the body, he explained.

Hawk is a member of the steering committee for the NCI-sponsored Adenoma Prevention with Celecoxib study, which stopped administering the drug in December 2004 after an independent committee discovered an increased cardiovascular risk among drug recipients.

He and an epidemiologist discussed the future for COX-2 inhibitors in cancer prevention at the American Society of Preventive Oncology annual meeting in San Francisco, held March 13–15.

A change in target patients

Hawk said NCI, despite the revelations about rofecoxib and celecoxib, has not abandoned the idea of using COX-2 inhibitors to reduce patients’ risk of colorectal cancer.

The Web site ClinicalTrials.gov showed in March that recruitment continued for an NCI-sponsored study of celecoxib, with or without oral eflornithine, in the prevention of colorectal cancer in patients with familial adenomatous polyposis.

Colorectal cancer develops in nearly all patients with familial adenomatous polyposis by age 40 years if they have not had the precancerous polyps surgically removed, according to the American Cancer Society.

“In high-risk cohorts,” Hawk said, “we’re attempting to continue on with the celecoxib trials because there we think . . . the potential benefits outweigh potential risks, and we’ll see if everyone shares our view.”

So far, he added, investigators and personnel at NCI and FDA “are on the same page with regard to the importance of moving that research forward.”

When NCI suspended use of the COX-2 inhibitor in the Adenoma Prevention with Celecoxib study, 86% of the 2035 patients had already taken the drug or placebo for the three years that had been planned, according to Hawk’s presentation in February to two FDA advisory committees.

All of the patients had at least one adenomatous polyp surgically removed from their colon or rectum before they started the study and had a moderate risk of colorectal adenoma, he said. Celecoxib’s ability to prevent the growth of new polyps, and hence reduce the risk of colorectal adenoma, would be assessed one and three years after the polypectomy.

Hawk said at the March meeting that the research team hopes to have the efficacy data in hand before the end of the year.

Also at the FDA-convened meeting in February, Merck Research Laboratories reported unpublished efficacy results from its Adenomatous Polyp Prevention on Vioxx study, whose cardiovascular data had prompted the company to stop marketing rofecoxib.

Daily therapy with rofecoxib 25 mg daily for three years had reduced the relative risk of colon-polyp recurrence by 24% in patients with colorectal adenoma, Merck’s Ned S. Braunstein had told FDA and its advisory committees.

As for the NCI-sponsored studies of COX-2 inhibitors in patients at lower risk of colorectal adenoma, “that line of research is not moving forward at the moment,” Hawk said.

NCI, he said, now focuses its chemoprevention efforts with COX-2 inhibitors on high-risk patients—those with a 40–100% lifetime risk of developing a specific cancer.

One goal of NCI’s continuing research is to identify patients with relatively low safety concerns from long-term exposure to COX-2 inhibitors who would benefit greatly from the reduction in cancer risk, he said.

Not the first surprise

“The trials of selective COX-2 inhibitors give us a sense of ß-carotene revisited,” said discussion moderator Michael Thun, head epidemiologist at the American Cancer Society, referring to nearly decade-old research.

The Beta-Carotene and Retinol Efficacy Trial, a large placebo-controlled study that involved patients at high risk of lung cancer, had found an increased risk of death in general and a possible increased risk of death from cardiovascular disease among supplement users. Use of the two supplements in the study stopped in early 1996 when the steering committee discovered the increased risks.

At about the same time, results of the ß-carotene component of the Physicians’ Health Study had been released. That portion of the study had found that long-term use of the supplement had no effect on the rate of death, cancer, or cardiovascular disease in the participants, half of whom were current or former smokers.

Both prospective studies of ß-carotene had been undertaken on the evidence of epidemiologic findings and laboratory investigations.

Breast cancer prevention

Twenty-one epidemiologic studies have focused on the occurrence of breast cancer and the use of nonsteroidal antiinflammatory drugs (NSAIDs), primarily aspirin, said Randall E. Harris, an epidemiologist and biostatistician at the Ohio State University in Columbus. On the basis of those studies, Harris’s research group estimated that the risk of breast cancer was reduced by 39% with the daily use of aspirin 325 mg or ibuprofen 200 mg—drugs that inhibit COX-2 and cyclooxygenase-1.

Harris directed the highly publicized Women’s Health Initiative study that found a 49% reduction in the relative risk of breast cancer with long-term use of ibuprofen, usually 200 mg three times a week, and a 21% reduction with aspirin therapy among postmenopausal women.

Preliminary findings from his university-conducted study of 300 women with breast cancer who used a pain reliever and 300 women without the diagnosis who used the same medication suggest a cancer-preventive effect from NSAID exposure, he said.

Whereas acetaminophen, which lacks an antiinflammatory effect, was associated with a 10% reduction in the relative risk of breast cancer, the NSAIDs in the study yielded better results, Harris reported. COX-2-selective NSAIDs, primarily rofecoxib and celecoxib, were associated with a 49% reduction in users’ relative risk, aspirin with a 34% reduction, and ibuprofen with a 74% reduction.

He said he was not surprised by these findings, given that the COX-2 gene and its enzyme affect cell division, the development of mutations, formation of new blood vessels, transmission of cancerous cells to secondary sites, and programmed cell death.

“I think we need to continue to explore the role of COX-2 inhibition and the combination of COX-2 inhibitors with perhaps other types of chemotherapeutic agents [in decreasing] the risk of not only cancer but also the risk of other chronic diseases,” he told the audience.

What matters more?

“The fundamental issue,” Thun said in describing what he sees as the new paradigm for drug development in cancer chemoprevention, “is that in the absence of safety, efficacy doesn’t matter.”

Safety, he added, is “especially difficult to achieve when the goal is to prevent a single type of cancer in healthy people.”

NCI’s Hawk disagreed with Thun’s paradigm.

“Clearly, safety is extremely important in [disease] prevention,” Hawk said. “Nevertheless, we wouldn’t recommend anyone take anything unless efficacy was demonstrated first. So in my view, efficacy does come first, and then safety comes second.”