United Press International - September 19, 2000

TORONTO, Sept. 19 (UPI) -- Regular, skilled physical breast examinations do as well as regular mammograms in preventing deaths from breast cancer, a controversial Canadian study shows.

      "There was no benefit in terms of reduction of the number of deaths by using mammography," lead investigator Anthony Miller of the University of Toronto said. The study was to be published Wednesday in the Journal of the National Cancer Institute.

      But a prominent critic of the study, radiologist Dan Kopans of Massachusetts General Hospital in Boston, said the result is "beyond belief" and criticized the journal for publishing it. When preliminary data was released in the early 1990s, Kopans raised questions about the study design and the way it was carried out. He told United Press International that he hasn't changed his mind.

      The Canadian National Breast Screening Study followed more than 39,000 Canadian women aged 50 to 59 for 13 years. They were split into two groups: one got regular physical breast exams combined with mammograms and the other only got regular physical exams.

      To the end of 1996, Miller said, 107 of the women in the mammography group had died of breast cancer compared with 105 in the physical exam group. The picture was similar for cases of invasive breast cancer: screening detected 622 cases in the mammography group, compared with 610 in the exam group.

      "This offers women a choice," Miller said, especially in places where high-quality mammography is not available. But he added that the physical exams used in the study are intensive, 10-minute tests that require a high level of training and may not be available in many places.

      He added that the study does not invalidate earlier studies that showed regular mammograms -- breast x-rays -- prevent cancer deaths. "This does not show that screening does not work," he said, but only that other forms of screening can work as well.

      "We know that mammography is effective," said Barbara Whylie, medical affairs director of the Canadian Cancer Society, which was one of the study's sponsors. "This study says that a properly done skilled breast examination can be as effective."

      But Whylie said the cancer society will not be changing its guidelines -- which call for regular mammograms for women over 50 -- because proper breast exams are not available to all women.

      Epidemiologist Cornelia Baines of the University of Toronto, one of the study's co-authors, said the usual one- or two-minute breast exam is not enough. "You're not looking for golf balls, lemons or oranges," she said. "You're looking for subtle signs."

      She said a good examination covers "every square centimeter" of the breast-from the armpit to the breastbone and from the collarbone to bottom of the ribcage. "If a woman gets an examination like that, she can be sure she's having a good examination," Baines said.

      Kopans was more emphatic in his criticism. He told UPI the study was flawed on several grounds, but a key factor was that the physical exams were "more intense clinical breast exams that anyone in the U.S. does" and required special training for the nurses who carried them out. At the same time, he said, the quality of the mammograms was inferior.

      "They basically compared the highest quality clinical breast examinations to the lowest quality mammography," he said.

      Because the high-quality physical exams are not widely available, he said, it's "unconscionable" for the Journal of the National Cancer Institute to publish the study, without at least a warning that its findings are controversial. But radiologist Dan Sullivan, associate director of the NCI's biomedical imaging program, said: "it's a good study" although not surprising in light of the preliminary results.

      Like Canada's Whylie, Sullivan said the NCI would not be changing its guidelines on cancer. "We think mammography is valuable and we continue to recommend that," he said. "But the study does suggest that women could have a choice."

      Kopans has also criticized the study for the way women were assigned to the two arms. Usually, he said, the assignment is carried out blind, but in the Canadian study, the women were given a physical exam before they were assigned to either group.

      "So their clinical status was known," he said, which opens the possibility that the study samples were skewed in favor of the physical exam group.

(C) 2000 UPI All Rights Reserved.