The new study results come from researchers at the University of California, San Francisco and the University of Michigan Health System. They used a computer model to compare the cost-effectiveness of colonoscopy and other colon cancer screening procedures with that of Cox-II inhibitors, a class of drugs currently used against arthritis but also considered promising in preventing colon cancer.
The study looked at both approaches alone and together, for people with an average risk of colon cancer, and for those with a high risk due to family history.
The data, which will be published in the May issue of the American Journal of Medicine, suggest that the drugs are unlikely to be as effective -- dollar-for-dollar -- as colonoscopy in cutting cancer death risk. The finding holds true both for those with average colon cancer risk, and for those with a higher family-linked risk, even if studies now under way around the country show the medications cut cancer risk significantly. Screening plus drugs was most effective, but even more costly.
"Rather than wait to see if a pill can protect you, it makes much more sense to follow the advice that's already out there: Get a colonoscopy every five to ten years according to your risk, because this relatively simple procedure can help save your life," says co-author James Scheiman, M.D., an associate professor of internal medicine at the U-M Medical School.
Scheiman adds that he and his colleagues aren't disputing the potential effectiveness of Cox-II inhibitors like celecoxib (sold as Celebrex) and rofecoxib (sold as Vioxx) in warding off polyps and cancer in the colon and other areas. But their finding suggests that the drugs would have to be extremely effective and inexpensive in order to beat the cost-effectiveness of colonoscopy.
"Colon cancer is the fourth most commonly diagnosed cancer and the second leading cause of cancer death in this country, killing 57,000 people a year, and we need every weapon against it that we can find," says co-author Uri Ladabaum, M.D., M.S., an assistant clinical professor of gastroenterology at UCSF. "Right now, screening is the best preventive weapon we have, and we need to ensure that more Americans take advantage of it."
Colonoscopy allows doctors to see and remove polyps and cancerous lesions in the colon with great accuracy, and can cut cancer incidence by 60 percent. So, guidelines suggest that most people have a colonoscopy every 10 years, or other screening tests more often. Those with higher risk, due to a family history of the disease, should have a colonoscopy every five years.
Medicare covers a colonoscopy every 10 years for all participants, in addition to coverage of other tests for all, and coverage of more frequent colonoscopy for those with increased risk.
But only a fraction of adults over 50 who should have a colonoscopy or other colon cancer screening regularly actually do. Partly as a result, only 37 percent of colon cancer cases are diagnosed while tumors are still confined to the local area, when treatment is most effective. In the face of this poor compliance with screening guidelines and high death toll, much attention has focused on promising evidence that Cox-II inhibitor drugs might be able to cut cancer risk.
In fact, it was the promise of an easy-to-take preventive measure that could supplement or even replace colon cancer screening that led Ladabaum and Scheiman to undertake the study with Mark Fendrick, M.D., associate professor of internal medicine at the U-M and co-director of the Consortium for Health Outcomes, Innovation and Cost-Effectiveness Studies, or CHOICES.
The team had previously analyzed the cost-effectiveness of colonoscopy, or a combination of sigmoidoscopy and stool testing for occult blood, compared with aspirin. They found that aspirin's toxicity was a major factor in keeping it from being a cost-effective alternative to screening. They published those results in the Annals of Internal Medicine in November, 2001 while Ladabaum was on the faculty at U-M.
Recently, the New England Journal of Medicine published two clinical research study results and an editorial on the same topic -- and both the studies found that aspirin had only a modest effect on preventing colon polyps. Each of these reports reiterated the main conclusion of the previous U-M study: that aspirin's effectiveness in reducing colorectal cancer risk will never be superior to the life-saving impact of colonoscopy.
Even as the aspirin clinical trials were under way, the U-M/UCSF team decided to compare the known costs and benefits of screening against the costs and benefits that might soon be established for Cox-II inhibitors once current studies are completed.
Cox-II inhibitors, which reduce inflammation and pain in the joints of people with arthritis, cut the production of an enzyme called cyclooxygenase-II, or Cox-II, which is made by cells in precancerous polyps and cancerous lesions. Because the drugs don't affect production of Cox-II's sister enzyme, Cox-I, they don't cause the stomach upset that other painkillers, like aspirin and ibuprofen, can create when they inhibit both kinds of cyclooxygenase.
Those two properties have made the drugs attractive for treating chronic joint disorders, and have raised hope that they could be effective against cancer. Already, the U.S. Food and Drug Administration has approved the use of high doses of celecoxib to prevent polyps in people with a rare disorder called familial adenomatous polyposis, or FAP.
Now, studies are proceeding to establish what effect the Cox-II drugs have on the development of polyps and cancerous lesions in people with other colon cancer risk factors or average risk. The studies will also help establish what doses of the drugs are needed to produce an effect.
Because those data aren't yet available, the new U-M/UCSF study used assumptions to compare the costs incurred and years of life gained using Cox-II inhibitors with those gained by colonoscopy screening and the two techniques combined. The computer model calculated results for people with various cancer risks over a 30-year period, starting at age 50.
The model assumed that people without a high risk of colon cancer would have a colonoscopy once every 10 years, unless a polyp was found, at which time screening would occur every 5 years. People with higher risk were modeled as being screened once every 5 or 10 years. And people taking a Cox-II inhibitor were modeled as taking the drug twice daily.
The study looked at the number of life-years saved by screening, drugs, or a combination -- that is, how many years of life a person would "get back" if their pre-cancerous polyps or tumors were detected early by a colonoscopy or prevented by the Cox-II drugs. They based the expectations of cancer risk and incidence on recent statistics. Then, they looked at the cost for each life-year saved, based on the cost of the screening, long-term drug treatment, or combination.
For example, the study found, if Cox-II inhibitors reduce colon cancer risk in the average person by 30 percent, at a dose that costs a dollar a day, the drugs will be far less cost-effective for average people than colonoscopy, which reduced cancer risk by 60 percent to 70 percent.
For an average-risk person, it would cost $20,200 to save one life-year through colonoscopy screening, but $233,300 to save the same life-year through Cox-II drugs.
Even for a person with a higher risk of colon cancer, the study found, a colonoscopy once every five years would still cost less and save more years of life than a daily drug. But the difference was smaller: $3,900 per life-year saved through screening every 10 years or $6,200 for screening every 5 years, versus $80,300 for drug therapy.
Though the study showed that the cost-effectiveness of colonoscopy far outweighed that of drug therapy or a combination of screening and drugs, the computer model did reveal that there might be certain conditions under which Cox-II inhibitors could match colonoscopy in preventing colon cancer.
"If the clinical trials now under way show that Cox-II inhibitors reduce colon cancer risk by the same percentage -- approximately 60 percent -- as colonoscopy, and the price of the drugs falls to 25 cents per day, the cost-effectiveness of the drugs may approach that of the screening for those with average cancer risk," says Ladabaum. "And the drugs would have to cut risk by 60 to 70 percent and be priced at 50 cents a day to be comparable to screening in those with moderately high familial risk."
But, says Fendrick, "The Cox-II inhibitor results reinforce what we have been saying all along: There is no currently available drug that comes close to matching the cancer reducing effect of colonoscopy."
Reference: American Journal of Medicine, 114:7.
Additional contact person, for University of California, San Francisco: Maureen McInaney, mmcinaney@pubaff.ucsf.edu, 415-476-2557