Stool-Based Colon Cancer Tests Vary Widely in Accuracy
Feb. 4 -- TUESDAY, Feb. 3 (HealthDay News) -- Newer versions of the stool-based colon cancer tests -- recommended annually for Americans over 50 -- vary widely in their ability to spot potential abnormalities, a German study finds.
Doctors at 20 gastroenterology practices in Germany used the tests for what is called fecal occult blood -- otherwise undetectable traces of blood that could come from a cancer or polyp in the intestine -- on more than 1,300 people and checked those results against findings from colonoscopy and an X-ray examination of the colon.
The study focused on the two types of tests for fecal occult blood, each of which uses a small stool sample placed on a card. The older test uses a chemical called guaiac, which can detect the "heme" part of the hemoglobin molecule in blood. However, the accuracy of the guaiac test can be compromised by the patient's diet -- for example, the recent consumption of red meat or even vitamin C supplements.
In 2000, a newer type of stool screen called the "immunochemical" test was developed. It doesn't spot blood directly, but rather picks up antibodies to the globin in the blood's hemoglobin molecule. This test is less easily "fooled" by diet, and soon after its introduction, the American Cancer Society recommended adding the immunochemical version to its screening guidelines, noting that the tests "are more patient-friendly and are likely to be equal or better in sensitivity and specificity" to the guaiac screen.
By now, a number of immunochemical tests have entered the marketplace. The researchers, from the German Cancer Research Center in Heidelberg, compared six tests against each other and against the older guaiac test. They reported their findings in the Feb. 3 issue of the Annals of Internal Medicine.
The team found a wide variance in diagnostic performance between the tests. The sensitivity for accurately spotting an adenoma (suspicious growth) ranged from 25 percent to 72 percent, while the tests' ability to accurately weed out "false positives" ranged from 70 percent to 97 percent.
