DURHAM, N.C. -- Many of the erroneous results from the cheapest and most common screening method for colon cancer are caused by bleeding above the colon, a Duke University Medical Center researcher has found. Such colon cancer screening is often recommended for all Americans over the age of 50.
The test utilizes guaiacum, a substance derived from tropical trees that reacts with substances in blood and turns blue.
Physicians have debated the effectiveness of the guaiac test, which detects the presence of occult, or hidden, blood in the feces. On one hand, several large studies have demonstrated that the test saves lives by finding cancers early. But many patients with a positive test, indicating blood in the stool sample, are found to have no signs of colon cancer. Such a result, termed a "false-positive" test, results in expensive and potentially harmful examination of the colon for many patients.
In a study published in the July 16 edition of the New England Journal of Medicine, Duke gastroenterologist Dr. Don Rockey believes he has found an important reason for the high false-positive rate: these tests detect bleeding from the stomach and upper GI tract as well as the colon. In fact, bleeding appeared to be detected at a higher rate in the upper intestines than in the colon.
"These findings should make us think more closely about the various tests we use for the detection of blood in the stools and the way we use them to determine if someone has colon cancer," Rockey said in an interview. "These tests are good, but not perfect."
The test involves placing a few drops of a special reagent onto a card containing guaiac and a small stool sample. A blue reaction indicates there is likely to be blood (which generally can not be seen by the patient and therefore is termed "occult" blood) in the stool and can be an indication of cancer in the colon, the final segment of the large intestine. The test is inexpensive, readily available and easy to do. In most cases, patients collect the sample at home, and send it through the mail to their physician or to a lab for testing.
The study involved nearly 250 people who tested positive for occult blood with the guaiac stool tests. Each of the patients then received procedures where physicians actually look for bleeding sites through scopes -- a colonoscopy for the colon, and an esophagogastroduodenoscopy for the GI tract above the upper small intestine, including the stomach.
In the study, the source of the bleeding was found to be a lesion in the upper GI tract in 71 cases, and the lesion was in the colon in 54 of the cases.
Although not part of the study, Rockey notes that there are other problems that further complicate the reliability of the tests, including diet. For that reason, patients are instructed not to eat red meat or such raw vegetables as horseradish, cauliflower or broccoli before collecting their samples.
Rockey said that newer stool screening techniques, based on different methodology are being developed and perfected and may be more specific for colon cancer. He and others are currently examining such newer options.
The debate among physicians is over how best to screen an aging population that exceeds 50 million in the United States. Without symptoms or a family history of colon cancer, physicians recommend screening for colon cancer beginning at the age 50 for both men and women.
The guaiac-based test for occult blood in the stool is effective, but imperfect. Moreover, Rockey said this is highly controversial area among physicians because if blood is found, the procedures used are expensive and entail some degree of risk, including death. For example, the costs for a colonoscopy ranges from $500 to $1,300, and an esophagogastroduodenoscopy study ranges from $300 to $800.
"The last word on this and other methods for colon cancer screeing is not in; further studies are clearly needed," Rockey said.
The patients in the study were examined at San Francisco General Hospital and the San Francisco VA Medical Center from 1994 - 1996, when Rockey was a faculty member at the University of California, San Francisco. Joining Rockey in the study, from Duke, was Linda Sander, M.P.H. Also involved, from U.C.S.F., were Drs. Johannes Koch, John Cello and Kenneth McQuaid.
The study was supported by U.C.S.F. and Duke.
Journal
New England Journal of Medicine