Colorectal cancer kills more Americans than any other cancer except lung cancer. But the death toll doesn't have to be as high as it is. Screening works. The American Cancer Society estimates that such tests saved 70,000 lives in the last 20 years.
"Just think how many we could save if we did it right," says Dr. Otis Brawley, chief medical officer for the ACS in Atlanta.
Doing it right would mean doing it on a lot more people. During those 20 years, only about 25% to 30% of the men and women who should have been screened — those age 50 or older — did get screened, Brawley says. Even today, that figure remains below 50%.
Doing screening right also means using the most effective screening test, and a consensus has developed over the last decade that the most effective test is colonoscopy, in which the entire large intestine is explored, and not an older method, sigmoidoscopy, which examines only part of the bowel.
Doctors began to favor the thoroughness of colonoscopies about 10 years ago. The preference for colonoscopies has grown so great, in fact, that doctors say it can be hard to get a sigmoidoscopy if you want one, especially in urban areas.
The shift occurred not because of definitive proof that colonoscopy was a better test — that didn't exist — but because simple logic implied that it was.
"We all thought it made sense," says Dr. Linda Rabeneck, a gastroenterologist who studies the effectiveness of colorectal cancer screening at the University of Toronto. "Colonoscopy ought to be better."
Now some experts, including Rabeneck, fear that such logic may have led them astray.
In a colonoscopy, a flexible tube carrying a tiny camera is inserted through the rectum and threaded 4 to 6 feet through the large intestine to look for cancer and polyps. When it finds them, doctors can thread small surgical instruments through the scope to biopsy or remove them.
The test is similar to a sigmoidoscopy, which used to be the primary screening test for colon cancer. Instead of carrying out a search-and-destroy mission through the entire length of the colon, a sigmoidoscopy is limited to the 1 to 2 feet on the left side that are easiest to reach. That makes it a quicker, easier and safer procedure that doesn't require anesthesia and is a lot less expensive.
The American College of Gastroenterology favors colonoscopy, but some other organizations, including the American Cancer Society and the U.S. Preventive Services Task Force, do not single out any particular test to screen for colon cancer.
To be clear, no one contends that colonoscopy isn't as good as sigmoidoscopy at finding colon cancer. The question is simply whether colonoscopy is a better test — and, if so, if it's enough better to justify the increased effort, expense and risk it entails.
"Asking for a sigmoidoscopy is reasonable," says Dr. Alfred Neugut, a cancer researcher at Columbia University's College of Physicians and Surgeons in New York. In a debate-provoking commentary published in July in the Journal of the American Medical Assn., Neugut and Dr. Benjamin Lebwohl, a gastroenterologist at Columbia, make the case that sigmoidoscopies may have been consigned to the dustbin of screening tests prematurely.
Rabeneck agrees. "People have very different views on what they will undergo as a screening test," she says. "Colonoscopy is not for everyone."
And the evidence that colonoscopy is a better test is just not there at the moment, Neugut and Lebwohl argue. In three recently published studies it performed no better than sigmoidoscopy. None of the studies were randomized trials — the gold standard of medical research — and none compared colonoscopy and sigmoidoscopy directly.
One study, led by Rabeneck, compared more than 10,000 patients who were diagnosed with colon cancer and subsequently died with more than 50,000 other people matched by age, gender, location and socioeconomic status. Researchers found that having a colonoscopy was associated with fewer deaths from colon cancer.
Another study analyzed data from more than 3,000 patients who had screening colonoscopies, and a third analyzed records for more than 45,000 patients in a database of physicians' billing claims. In these two studies, having a colonoscopy was associated with a lower incidence of colon cancer.
In all three studies, the associations between colonoscopies and cancer or death were found only for growths on the left side of the colon — that is, growths that sigmoidoscopy would have found too.
Although these results were unexpected, they're not inexplicable. It's harder for a patient to prepare for a colonoscopy than for a sigmoidoscopy, since it requires cleaning out the entire bowel instead of just the lower section. It's also harder for a doctor to perform a colonoscopy, since the tube has to be maneuvered much farther and through more twists and turns. As a result, the quality of colonoscopies can vary a great deal.
In addition, the biology of the colon is different on the two sides. Polyps on the left — which are reachable with a sigmoidoscope — have a stalk and look a bit like mushrooms, while polyps on the right are generally quite flat. This makes them harder to find and snip out.
But just because the test has its complications doesn't mean it isn't — still — a better test, says Dr. David Johnson, chief of gastroenterology at Eastern Virginia Medical School in Norfolk and past president of the American College of Gastroenterology.
"It would be a gross overreaction and inappropriate response to infer that colonoscopy does not provide a benefit," he says. "That would be taking us back 20 years, at a time when we're making great strides in reducing colon cancer mortality."
Others are skeptical that colonoscopy will ever prove to be more valuable than sigmoidoscopy.
"If we had very, very, very high-quality colonoscopies, we still don't know if we'd see a benefit on the right side," Rabeneck says.
The debate is likely to continue for some time, perhaps until results come in from two large randomized trials in Europe that aim to nail down the benefit of colonoscopy by comparing patients who get them to patients who get other types of screening tests or no screening at all. But those studies, which each involve more than 50,000 patients, will take at least a decade to complete.
In the meantime, experts on both sides of the issue emphasize that either test is much better than no test at all. In fact, they say, any of the available screening tests is worthwhile.
Take fecal occult blood tests, which check for blood in stool samples. Although they've been overlooked somewhat in the current debate, so far they're the only other tests, besides sigmoidoscopies, to have been proved effective in a randomized trial.
Fecal blood tests are relatively low-tech. They're also relatively cheap, convenient and non-risky. And some doctors and researchers believe they don't get enough respect.
"I could do more for the public good, if I wereczar, if I started doing stool blood tests on everybody tomorrow, following up with colonoscopies if need be," says Brawley of the American Cancer Society. "That would save more lives than what we're doing now."
Still, for Brawley, every test is a favorite test: "The best test is the test you decide to get."