New test detects cancer of colon

Hopkins process finds earliest stages using gene analysis

January 31, 2002|By Jonathan Bor | Jonathan Bor,SUN STAFF

Scientists at the Johns Hopkins Kimmel Cancer Center are reporting that they have developed a simple genetic test that can detect a large percentage of colorectal cancers at their earliest stage.

The test, which requires patients to submit only a stool sample, finds a genetic error that appears years and sometimes decades before the cancer turns deadly.

"Deaths from colon cancer are nearly completely preventable if the tumors are detected early," said Dr. Bert Vogelstein, the pioneering cancer researcher whose lab has pursued the test for almost 10 years. It was devised to detect the most common form of colon cancer, which is not inherited.

The test, which could be ready for wide-scale use in three to five years, could mean far fewer people would have to undergo a colonoscopy, an invasive and more expensive screening tool that is performed under anesthesia, he said.

People who test negative in the genetic analysis would not need a colonoscopy; those who test positive would. That colonoscopy would be used to locate and possibly treat the tumor.

The study, published in today's New England Journal of Medicine, was done in collaboration with scientists at the M.D. Anderson Cancer Center in Houston, Uppsala University in Sweden, the Lahey Clinic in Boston, and Exact Sciences Corp. of Massachusetts.

Colorectal cancer is the fourth-most-common and second-deadliest cancer in the United States, with 135,000 cases diagnosed and 56,000 deaths occurring each year. Only lung cancer kills more.

In the Hopkins study, scientists analyzed stool samples from 74 people - 28 with early colon cancers, 18 with precancerous growths and 28 with no disease. The cancer patients had already received diagnoses through conventional methods.

Using the new test, scientists detected a genetic mutation in 61 percent of the people with early cancers and half of those with precancerous lesions. The test did not produce positive results in any of the healthy people, an encouraging sign that the test was not prone to the "false positives" common to other diagnostic tests.

"We still have a long way to go before we can confidently use such a screening test in the general population," said Dr. Kenneth Kinzler, who is involved in the research at Hopkins. "But we are encouraged by the fact that we've detected mutations in a significant fraction of the patients with early tumors and never in people free of disease."

Vogelstein said he believes the test's effectiveness could be raised to 70 percent with minor refinements. The detection rate could be raised further by combining the test with another that would flag a second mutation, he said.

But already, the detection rates are comparable to those achieved by mammography and Pap smears, which hover in the 60 percent to 80 percent range, he said. And those tests, he said, produce many false positives.

The test looks for alterations in the APC gene, a tumor-suppressor gene that is part of the body's natural defense against cancer when it functions properly. The gene defect is the first of four or five - all discovered in Vogelstein's lab in past years - that must occur for colon cells to turn cancerous.

Scientists at Hopkins have been working on the test for about 10 years. It is based on the observation that any developing tumor will shed cells into stool, which can be analyzed for telltale mutations.

But the work was hampered by daunting challenges. First, most of the DNA found in stool are not from human cells but from bacteria that populate the intestines.

Second, colon cancer is triggered not by just one mutation, but by many that can occur at different places along the gene. And third, most of the APC genes shed into stool are normal; only about one in 100 are mutated in the cancer patient.

To overcome the problems, the scientists developed a technology they called "digital protein truncation" that makes the mutations easier to find.

Scientists decided to target the APC gene because the mutation typically occurs in a person's thirties or forties, about 20 to 30 years before the person develops full-blown cancer.

Tests to diagnose cancers from readily obtained body secretions are being eagerly sought both inside and outside Hopkins.

Dr. David Sidransky, a Hopkins researcher, is developing tests to screen saliva for oral cancer, and blood for evidence of head and neck, stomach and colon cancer.

A group at the Mayo Clinic published a stool test for colon cancer. The scientists achieved a higher detection rate but turned up many false-positive cases.

Dr. Andrew C. von Eschenbach, who will be installed next week as director of the National Cancer Institute, said the Hopkins study represents one of the ways scientists are using knowledge of cancer genetics to diagnose and, potentially, to treat the disease.

"It's a great example of this incredibly powerful story," he said.

Dr. Barbara Conley, chief of the institute's diagnostic research branch, said the detection rate will have to be improved before patients can safely pass up colonoscopy.

"A negative test then doesn't mean you don't have something," she said, noting that the screen failed to find cancers in 40 percent of the people who had the disease.

Baltimore Sun Articles
Please note the green-lined linked article text has been applied commercially without any involvement from our newsroom editors, reporters or any other editorial staff.