Roanoke Times
                 Copyright (c) 1995, Landmark Communications, Inc.

DATE: THURSDAY, May 13, 1993                   TAG: 9305130064
SECTION: NATIONAL/INTERNATIONAL                    PAGE: A-5   EDITION: METRO 
SOURCE: Associated Press
DATELINE: BOSTON                                LENGTH: Medium

COLON TEST MAY LOWER DEATHS BY A THIRD

The study is the second major development this month in colon cancer research. Last week, scientists reported the discovery of a flawed gene that is responsible for about one in seven of cases of colon cancer.

Researchers who conducted the latest study estimate that giving the $5 fecal blood test to everyone over age 50 could prevent 20,000 deaths annually from colon cancer, the nation's second leading cancer killer after lung cancer.

"We are the first study to have a conclusive result to show that the test is effective for reducing mortality," said Dr. Jack S. Mandel, who directed the research at the University of Minnesota.

However, experts differed over whether the study is the proof needed to urge everyone over age 50 to get the test yearly, as the American Cancer Society recommends.

At issue is the test's accuracy in diagnosing colon cancer. Ominous results almost always turn out to be false alarms.

Overall, 10 percent of those tested are found to have blood in their stools, a possible sign of colon cancer. Less than three in 100 actually have colon cancer. Yet the 97 others must undergo costly, unpleasant follow-up tests to prove they are cancer free.

Because of the high cost of the follow-up tests to rule out cancer, mass screening of older Americans could cost more than $3 billion a year, according to one estimate.

"It's not good enough to apply on a mass screening basis," contended Dr. Douglas B. McGill of the Mayo Clinic.

However, Dr. William Carey of the Cleveland Clinic, president-elect of the American College of Gastroenterology, commented that while the test is imperfect, the latest study "demonstrates that doing something is better than doing nothing."

The study, begun in the mid-1970s, was conducted on 46,551 people ages 50 to 80. They were randomly assigned to have the test annually, every other year, or not at all.

During 13 years of follow-up, six of every 1,000 people tested died from colon cancer, compared with nine of every 1,000 who did not get tested. Those tested every other year had only a slightly reduced death rate.

The study, financed by the National Cancer Institute, employed the Hemocult test, which is widely used in doctors' offices and at home. A stool sample is smeared on a piece of treated paper, which is then usually sent off to a lab to be checked for blood. However, blood in the stool can have many possible sources, including ulcers and eating red meat.

Typically, people with positive tests are checked for cancer with colonoscopy, a test that involves inserting a flexible tube through the rectum into the colon. Colonoscopy typically costs between $500 and $1,000.