Wednesday, July 02, 2003

By Anita Srikameswaran, Post-Gazette Staff Writer

A large national study led by local researchers indicates that screening for colorectal cancer every three years instead of the recommended five years can uncover a small number of cancers that might otherwise have gone undetected -- and untreated -- for another two years.

But the results are not enough to prompt a change in screening guidelines, which recommend flexible sigmoidoscopy every five years for people 50 and older. In the procedure, the lower portion of the large intestine is examined with a flexible, fiber optic scope.

Of more than 9,000 study participants, six were found to have a cancer and 72 to have an advanced, precancerous lesion three years after their first sigmoidoscopy. The findings were published today in the Journal of the American Medical Association.

Although there weren't many new cancers detected, the study raises questions about how frequently screening exams, including a more extensive exam called a colonoscopy, should be done, said lead investigator Dr. Robert Schoen, a gastroenterologist at the University of Pittsburgh School of Medicine.

"It's not a large number of cancers, on the one hand," he said. "On the other hand, it does show that the more frequently one tests, one will find things that are potentially of importance."

The study is by the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial Group, a federally funded research project that has recruited more than 154,000 people from 10 medical centers to see whether various screening tests reduce the number of deaths from those cancers.

For the colorectal cancer study, 9,317 people who had a sigmoidoscopy and had no sign of cancer or precancer underwent the same procedure three years later.

Almost 14 percent, or 1,292 patients, had an abnormal follow-up exam, meaning a polyp or mass was detected. Of those, 847 then had a colonoscopy, in which the entire large intestine was examined with a scope, and 104 had another sigmoidoscopy. In the re-examined group, 72, or 25 percent, had an advanced precancerous polyp and 6, or 0.6 percent, had cancer in the sigmoid colon.

It's not known whether the suspicious polyps might have been present during the first examination, Schoen said.

"There's no way to know whether it was missed or whether it grew [later]," he said. But 80 percent had advanced precancerous lesions in a portion of the colon that was "adequately examined at the initial sigmoidoscopy," the researchers said.

Many experts contend that colonoscopy, which is recommended every 10 years after age 50, is the better screening test because it can be done less often and the entire colon is examined. Studies indicate that 1 to 2 percent of people will have an advanced polyp or cancer detected by colonoscopy that would not be found with sigmoidoscopy, the researchers said.

Schoen said that in light of the new findings, a 10-year interval between colonoscopies may be too long. So far, no one has completed a study that looked at the number of dangerous lesions that might be found if colonoscopies were done more frequently.

It might be necessary to do sigmoidoscopy or other screening tests between colonoscopies, Schoen said. Better yet, blood tests that can detect problems need to be developed.

For now, "I don't think we should change the recommendation from five years to three years," he said. "We need to watch this very carefully and we need more studies."

Dr. Robert Fletcher, of Harvard Medical School, came to the same conclusion.

"All screening tests, sigmoidoscopy included, are imperfect," he wrote in a commentary also published today. "Screening tests are chosen because they are easier, safer and less expensive than definitive tests, but at the cost of missing some cancers and sounding many false alarms."

Fletcher noted that it's not known whether finding the advanced precancerous polyps will prevent colorectal cancer deaths, the main goal of the trial, "because it is not known how often [they] progress to symptomatic, incurable cancers."

Schoen said that another arm of the study is examining a group of people who had sigmoidoscopies five years apart.

Anita Srikameswaran can be reached at anitas@post-gazette.com or 412-263-3858.