The Screen Team

As tumors go, those in the colon and rectum are among the most preventable. In their early stages, they’re also beatable. Yet every year in the United States, nearly 150,000 new cases of colorectal cancer emerge, and the disease kills about 55,000 people. Those numbers make colorectal cancer the fourth-most-common and second-most-lethal malignancy. The problem, doctors say, is that many people don’t get screened for the cancer when they should.

Screening is recommended for people age 50 and up. Optical colonoscopy, the most thorough test, can alert doctors to an emerging threat—a precancerous growth called a polyp—months or years before it would turn malignant. Polyps often form lobes that protrude into the hollow space between the colon’s walls.

Screening can also reveal a colon tumor before symptoms appear. At that point, the usually slow-growing cancer can be removed surgically.

“If you catch it early, it’s nearly 100 percent curable,” says gastroenterologist Steven Itzkowitz of Mt. Sinai School of Medicine in New York City.

However, according to a national survey in 2000, less than half of people over 50 have been adequately screened. Some people get a colonoscopy once and never get another.

“We are just not that successful at getting patients in for screening,” Itzkowitz says.

Colonoscopy is time-consuming and unpleasant. The day before testing, patients must consume only liquids and must empty their bowels completely using medicines that sometimes cause nausea. Patients must then be sedated during the exam, and about 1 exam per 500 results in a serious complication, such as puncturing of the colon.

During a colonoscopy, doctors insert a thin, lighted tube into a patient’s rectum to examine the length of the colon. If they find a polyp, they remove it by, for example, severing it with a wire loop. They can also take tissue samples to biopsy for cancer.

The recommended interval between colonoscopies is 10 years for most people. In contrast, doctors recommend that flexible sigmoidoscopy, which uses a similar method to check a smaller portion of the colon, be repeated every 5 years.

Less invasive tests, such as fecal-blood tests and barium enemas, provide less information about a patient’s cancer risk than colonoscopies and sigmoidoscopies do, and they must be repeated more frequently.

Alternative screening methods could take a bite out of colon cancer’s toll. Medical researchers are devising new, less invasive or noninvasive tests that may make screening more appealing. An X-ray technique called virtual colonoscopy is already in limited use. Some other methods now being tested analyze a sample of blood or feces.

But the new tests must be shown to be effective in identifying, among millions of people, those who have colorectal cancer or are at high risk of developing it.

Elusive cancers

Prompt screening could have helped Katy Duggan. In November 2000, at age 50, the Seattle resident was diagnosed with colon cancer. In retrospect, she says, hints of the disease had showed up several months earlier. Streaks of bright-red blood appeared in her stool, which became progressively narrower as a tumor grew to block her colon. She also developed anemia, fatigue, and constipation.

But Duggan, a nurse, shrugged off those clues, dismissing them as symptoms of hemorrhoids, her unhealthy diet, and her demanding schedule. Eventually, the tumor obstructed her colon, leaving her completely constipated. Two weeks later, emergency room X rays identified the tumor.

Duggan survived her battle with cancer, but “it took one year out of my life,” she says. Chemotherapy, which lasted 6 months, sapped her of energy and creativity, and she became unemployed for a time.

Recently, doctors at Virginia Mason Medical Center in Seattle told Duggan that she is free of the disease. They recommend that she get a colonoscopy twice a decade from now on.

When colonoscopy first came into use 37 years ago, it was used as a follow-up test after other screening methods yielded abnormal results. Since then, “colonoscopy has become established as one of the preferred modalities for colon cancer screening,” says gastroenterologist Harminder Singh of the University of Manitoba in Winnipeg. However, he notes, it’s still fallible.

In a recent study, Singh and his colleagues tracked more than 32,000 people who had had colonoscopies that showed no evidence of colorectal cancer or precursor polyps. They found that 113 people were diagnosed with colorectal cancer within 5 years of screening.

Since past studies suggest that it takes more than 5 years for the smallest polyps identifiable by colonoscopy to develop into cancers, the screening tests of the patients in Singh’s study probably missed signs of danger, his team concludes in the May 24/31 Journal of the American Medical Association (JAMA).

Some polyps or cancers may have been missed because the patients didn’t maintain a liquid diet in advance of the test or didn’t drink all of the bowel-emptying preparatory medication. “If the bowel is not clean, colonoscopy can miss lesions,” Singh says.

Another factor contributing to the missed diagnoses may be a failure of the doctors to fully inspect difficult-to-access reaches of the colon. Singh’s findings suggest that doctors most frequently miss growths in the right colon, which is toward the top of the gastrointestinal tract. “We need to look in every nook and corner,” he says.

The JAMA study also posed a crucial question, says Singh: “Once colonoscopy has been done, how often should it be repeated?”

The rate of colorectal cancer among people who’d been screened 10 years earlier and found to be polypfree was only about one-quarter as high as it was in people who hadn’t been screened, his team found. That shows that current recommendations are reasonable, Singh says.

Soon after Duggan’s diagnosis, her two brothers and sister got colonoscopies. One brother had two polyps, which his doctor removed. But their mother, Norene Duggan, declined to be screened.

“I just have never felt that I needed it,” says the 89-year-old Griswold, Iowa, resident.

She may be right to sidestep the procedure today. Because malignancies in the colon progress slowly, older people “may die from other causes before a polyp has any chance of developing into cancer,” says Otto Lin, a gastroenterologist at Virginia Mason.

He and his collaborators considered 1,244 colonoscopies in people in three age groups. The team calculated that each colonoscopy administered in the 50-to-54-year-old group saved, on average, 10 months of life.

By contrast, in the group of 75-to-79-year-olds, a colonoscopy added only 2 months to life expectancy. For people over 80 years old, the life extension was barely 1.5 months. The study appears in the May 24/31 JAMA.

Although there were no major complications in the study, other work has shown that older people have elevated rates of colonoscopy-related problems, such as accidental perforations of the bowel. Whether patients older than 75 or 80 should have colonoscopies, Lin says, requires a judgment call that these people need to make with their doctors.

Because less invasive alternatives to colonoscopy are potentially safer as well as more convenient, they might be particularly attractive to older patients, he adds.

Going virtual

A well-studied alternative to colonoscopy is called virtual colonoscopy. For that test, technically known as computed tomographic (CT) colonography, doctors use X rays and computer programs to visualize the colon’s interior shape.

In preparation, patients drink bowel-emptying fluids just as they do for a colonoscopy. Carbon dioxide gas is then pumped in to inflate the bowels and make any polyps stand out. There’s no need to insert a scope into the bowel, so there’s also no need for sedation. The gas can, however, produce a bloated feeling.

Compared with standard colonoscopy, the virtual test is “just as good or slightly better” at detection of important polyps and tumors, says radiologist Perry Pickhardt of the University of Wisconsin Medical School in Madison. “Of course, it can’t remove polyps, so you still need optical colonoscopy for therapy,” he says.

In a recent unpublished study, Pickhardt found that follow-up colonoscopy is needed after fewer than 10 percent of virtual colonoscopies.

In an earlier study of 1,233 people, virtual colonoscopy identified 93.8 percent of the most dangerous polyps, those 10 millimeters or more in diameter. That compared favorably with the 87.5 percent of such lesions found by standard colonoscopy, Pickhardt and his collaborators reported in 2003. The Food and Drug Administration approved the screening procedure the following year.

However, in the trial, virtual colonoscopy slightly underperformed against colonoscopy in identifying polyps as small as 6 mm.

“Virtual colonoscopy is known to have a blind spot when it comes to spotting smaller polyps,” comments Itzkowitz.

While most evidence suggests that virtual colonoscopy is comparable to optical colonoscopy for detecting large polyps, not all data agree. For example, a 2004 study using an experimental method of virtual colonoscopy found that the test spotted only 55 percent of 10-mm polyps. However, to interpret the X-ray results, doctors in that study used methods and software that differed from those that Pickhardt’s group had employed.

Virtual colonoscopy is more expensive than the optical method and usually not covered by insurance, Itzkowitz says.

While noninvasive, virtual colonoscopy can’t be called pleasant. In the 2004 trial, volunteers who underwent both forms of screening expressed no consistent preference for one or the other.

A more tolerable test, says Itzkowitz, “might make it easier for patients to accept the idea of screening.”

A menu of options

Several teams are looking for tests of samples that are easy to collect. Fecal testing, which searches for traces of blood in the stool, has a history as a diagnostic tool for colorectal cancer. Before colonoscopy was developed, fecal occult-blood testing (FOBT) was the workhorse of screening methods. But that approach is far from perfect.

“Not all bleeding is cancer, and not all cancers bleed,” Itzkowitz says. Consequently, the occult-blood tests can misdiagnose people who don’t have cancer and, more importantly, falsely reassure some people who do.

“Well under half of [colorectal] cancers are picked up by fecal occult-blood testing,” Itzkowitz says.

A new fecal-DNA test, by contrast, looks for human-genetic material that contains certain abnormalities associated with colorectal cancer. Tumors slough off this DNA into stools. However, the DNA degrades quickly, so a recent clinical trial of the test asked 5,500 seemingly healthy volunteers to put fresh stool samples on ice and express ship them to a central lab.

Itzkowitz and his colleagues then compared the DNA-test results with those of FOBT and colonoscopy.

“Even though we express-couriered the stools and had ice packs around them, the DNA got degraded in transit,” Itzkowitz says. That complicated the laboratory’s task of picking up signs of cancer.

“The DNA test was four times better at finding cancers than the FOBT,” Itzkowitz says. Still, the DNA test missed 15 of the 31 tumors later identified by colonoscopy. “We felt that the test should have done better than a 52 percent pickup rate for cancer,” Itzkowitz says.

Exact Sciences Corp. of Marlborough, Mass., supported the clinical trial, which was reported in 2004.

Since then, Itzkowitz and his collaborators have enhanced the DNA test in three ways. They added a DNA-preserving buffer solution to the collection kit. They also improved their techniques for recovering traces of abnormal DNA from stool. Finally, they identified a new cancer-associated mutation and added an assay for it to the battery of 21 other genetic markers that the test examines.

A subsequent study auditioned the revised fecal-DNA test in 122 healthy volunteers and 40 people who had colorectal cancer. The test identified 88 percent of the tumors but deemed suspicious 18 percent of the stools that came from healthy people, Itzkowitz reported at Digestive Disease Week in Los Angeles in May. An abnormal finding would require colonoscopy to verify the presence of cancer, he says.

An intermediate version of the test, with two of the recent enhancements, has been commercialized. Like FOBT, the DNA test isn’t designed to detect polyps, and like virtual colonoscopy, it isn’t routinely covered by health insurers. But given the DNA test’s latest results, says Itzkowitz, “it ought to be considered an option for patients, especially those who are reluctant to get a colonoscopy.”

Other noninvasive tests are in earlier stages of development. For example, Given Imaging of Yoqneam, Israel, makes swallowable, camera-containing capsules that can reveal problems in the small bowel or esophagus. At meetings this fall, researchers affiliated with the company plan to present the first colon cancer-screening data collected by such devices.

The Johns Hopkins University research team that devised the fecal-DNA test and licensed it to Exact Sciences is now working on a blood test. It, too, searches for telltale DNA that cancers release.

Since blood is drawn during many routine medical appointments, screening blood for colorectal cancer—and potentially for other cancers—would require no additional action on the patient’s part, says Bert Vogelstein of Johns Hopkins.

In a pilot study, his team analyzed blood samples from 10 healthy people, 11 people with precancerous polyps, and 22 people with colorectal tumors of various stages. The results appeared in the Nov. 8, 2005 Proceedings of the National Academy of Sciences.

“All the patients with advanced cancers could be detected easily. About two-thirds of the early cancers”—those treatable by surgery—”could be detected,” Vogelstein says.

However, the experimental test flagged just 1 of the 11 polyps. If it can be used only to spot mature threats, and not to nip polyps in the bud, then DNA-based screening would need to be repeated with more frequency than optical colonoscopy, Vogelstein says.

“The purpose is not to replace colonoscopy,” he says. “The idea is to offer options … to give the patient a kind of menu of tests.”

He conjectures, “People who are older are more likely to opt for a noninvasive test.”

That applies to Norene Duggan. When she learned that a blood test for colorectal cancer was under development, she warmed to the idea of being screened.

“How simple that would be, rather than drink all that stuff and spend the night in the bathroom,” she says. “Since I have my blood done every year, it would be very simple to do it.”