Annual Mammogram Screening Urged for Women in 40s

WASHINGTON—“There is really no reason to doubt that mammography screening works for women in their 40s,” and saves lives, said Stephen A. Feig, MD, director of the Breast Imaging Center at Thomas Jefferson Hospital and professor of radiology at the Jefferson Medical College, both in Philadelphia.

Addressing a plenary session of the 28th National Conference on Breast Cancer of the American College of Radiology, he cited “scientific proof from randomized trials” that mammographic screening cuts breast cancer deaths in younger women, and made a strong case for annual rather than less frequent screening of women in their 40s.

Randomized trials, he said, are “easier said than done.” However, most randomized trials have shown benefits for screening women in their 40s, he said, and the few that have not are marred by substantial flaws in their design and implementation.

The very influential HIP study done in New York in the 1960s established a significant reduction in breast cancer mortality with mammographic screening in a study population aged 40 to 64, Dr. Feig said. When broken down by decade, the results showed a 23% mortality reduction in women aged 40 to 49 with screening, but this was not statistically significant. Yet, he explained, “some of the other older decade groups did not reach statistical significance either.”

He finds it “amazing that this was overlooked,” and emphasized that “no one has ever said, don’t screen women in their 50s, because of this study’s findings.” In fact, he stated, the study enrolled too few women in their 40s to establish statistical significance.

The Great Guidelines Controversy

In 1993, the National Cancer Institute rescinded a guideline recommending mammography screening for women in their 40s, a decision made, in part, because of the findings of the National Breast Cancer Screening Study of Canada, a trial marred by many serious flaws, Dr. Feig said. Poor quality mammography and faulty randomization render this study’s results worthless, he maintains.

Only a third of the mammograms in the Canadian study met acceptable technical standards, he said. Further, a decision to do clinical exams before randomization encouraged coding clerks to place the bulk of those with palpable masses, including late-stage breast cancers, in the experimental group that would receive mammographic screening, in hopes that these women’s cancers would be found and treated. This rendered the results invalid, he explained.

Because breast cancer occurs less commonly in younger than in older women, studies of younger women need more subjects in order to find sufficient cases for statistical significance, he said. Younger women’s denser breasts also make good mammography technique more important. Furthermore, because many cancers in younger women grow faster than those in older women, the screening intervals must actually be shorter to avoid missing interval cancers.

Also influencing the NCI’s 1993 decision, Dr. Feig said, was the fact that metaanalyses done at that time showed no benefit for women in their 40s, although subsequent metaanalyses with longer term follow-up have shown a statistically significant benefit, with the mortality differences increasing and the confidence limits narrowing with each study.

In addition, metaanalyses of five trials done in Sweden found a statistically significant mortality drop of 29%. “Several of the Swedish trials by themselves have now attained statistical significance and have shown 35% and 45% reductions in breast cancer deaths for women in their 40s,” Dr. Feig observed.

The NCI had set statistical significance as the goal for recommending mammo-grams for women in their 40s, but when that evidence emerged, “they pulled the goal post back,” he said. The 1993 conclusion was “obsolete” even when made and is all the more so now.

“We were really incredulous” at the decision, he said, adding that Richard Klausner, MD, director of the National Cancer Institute, expressed surprise and disappointment, and referred the question to the National Cancer Advisory Board, which came back with a decision opposite that of the NCI.

The guideline was changed in 1997, Dr. Feig noted, and a number of leading bodies in addition to the NCI now recommend mammography screening for women in their 40s. Even two of the experts who served on the initial NCI panel now also advise annual screening for this age group, he said.

Dr. Feig also said that women in their 40s need to be screened every year rather than less frequently. Although none of the trials compared different screening intervals, he said there is compelling evidence that annual mammographic screening will save more lives than biennial screening.

Since breast cancers grow at different rates, some can reach the clinical threshold between biennial screenings, and the “lead time” to the clinical stage “will be shorter in younger women,” he said. “Therefore, we may have to screen younger women much more frequently than every 2 years. The incremental benefit of more frequent screening is greater in younger women because the cancer is growing faster.”

Indeed, he said, “it’s more than coincidental that the two trials with the shortest screening intervals showed the greatest impact on death rate.” In the Swedish Two County trial, younger women had proportionately twice as many interval cancers as older women. Interval cancers have lower survival rates than screen-detected cancers at all ages, he added.

Clearly, Dr. Feig concluded, younger women need more frequent mammography screening than they received in most trials in order to reduce the number of interval cancers.