New Questions on Mammography Screening for Younger Women

LEEDS, England, Dec. 8 -- Once again the question of whether women ages 40 to 49 should have routine screening mammograms has arisen.

Findings from the AGE trial, reported by Sue Moss, Ph.D., of the University of Leeds, and colleagues, in the Dec. 9 issue of the Lancet, cast doubt on whether the benefits of screening women under age 50 outweigh the potential harm.

The randomized trial of nearly 17,000 women in England, Scotland, and Wales, which followed women for a decade, called it unclear as to whether younger women will benefit from mammography screening.

The uncertainty comes as a growing consensus has emerged in the U.S. that, on balance, mammography screening every year or two for women 40 to 49 is a good idea.

Although much debated for many years, the National Cancer Institute, U.S. Preventive Health Services Task Force, and the American Cancer Society all now come down on the side of screening for younger women.

The USPHS task force hedges a bit, though. "The precise age at which the benefits from screening mammography justify the potential harms is a subjective judgment and should take into account patient preferences," says the task force. "Clinicians should inform women about the potential benefits (reduced chance of dying from breast cancer), potential harms (e.g., false-positive results, unnecessary biopsies), and limitations of the test that apply to women their age. Clinicians should tell women that the balance of benefits and potential harms of mammography improves with increasing age for women between the ages of 40 and 70."

In the AGE trial, 160,921 women, ages 39 to 41, were randomly assigned from 1991 to 1997in the ratio 1:2 to an intervention group of annual mammography to age 48 years or to a control group of usual medical care.

At a mean follow-up of 10.7 years the researchers found a reduction in breast-cancer mortality in the intervention group of 17%, in relative and absolute terms, which, however, did not reach statistical significance (relative risk 0.83 [95% CI 0.66-1.04], P=0.11). The absolute risk reduction was 0.40 per 1,000 women invited to screening (CI -0.07 to 0.87).

After adjustment for non-attendance at the first screen, the researchers reported an estimated 24% reduction in breast cancer mortality (RR 0.76, 95% CI 0.51-1.01).

The number needed to screen to prevent one death over 10 years was 2,512 (CI 1149-13,544) for seven to nine years, the researchers calculated. In practice, they said, women in the intervention group received an average of seven mammography invitations. The estimate, 2,315, was more favorable when the needed-to- screen analysis was restricted to women with the potential for 10 years' follow-up and breast cancer deaths within 10 years CI 1,059-12,495).

Although the reduction in breast-cancer mortality observed in this trial was not significant, it is consistent with results of other trials of mammography in this age-group, the investigators said.

The absolute benefit of screening in this age group in terms of deaths prevented will be lower than the benefit in older women, but the life-years saved per death prevented will tend to be greater, the researchers said.

Assuming an average of 35 life-years saved per death prevented, and a cost of about per woman invited to have a mammogram, the results to date suggest a yearly cost per life saved of approximately ,000, based on seven invitations per woman, the researcher calculated. However, they said, that the full effects of screening have not yet emerged, whereas the full costs are likely to be reduced in the future.

One potential problem with the study was the decision to use only single-view mammography after the initial screen, partly because of concerns about the effect of radiation in this age group. The National Health Service, the investigators said, now routinely uses two-view mammography for all screens, and this approach has resulted in improved detection rates and reduced recall rates.

Summing up, Dr. Moss said that although the reduction in breast cancer mortality in this trial is not significant, it is consistent with results of other trials of mammography alone. Results from Swedish trials, which were somewhat similar, showed a 15% reduction for women ages 40 to 44. These results, together with the AGE study lend support to a reduction in breast cancer mortality of 15% to 17%.

Further follow-up of the Age trial will provide more information about the full effect of screening in this age group, Dr. Moss said. "There is need for research to identify, perhaps by modeling, the benefit of commencing screening at different ages below 50 years. Costs, both financial and in terms of false positives, will be higher than with later screening in view of the fact that the absolute risk reduction will be less in younger women."

"Meanwhile," Dr. Moss added, "it is important that individual women are provided with full information about both the possible harms and costs of screening."

In a commentary in the same issue, Benjamin Djulbegovic, M.D., of the University of South Florida in Tampa and Gary Lyman, M.D., of the University of Rochester (N.Y.), assessed the AGE study and nine other randomized studies, which altogether enrolled nearly 400,000 women. They concluded that there was little change in estimated overall survival for women who started screening in the 40 to 50 age range. In addition, they said mammography is generally only about 85% accurate.

In a review of the trials, they found that women, ages 40 to 49 who were invited for screening received an earlier diagnosis of breast cancer than those not screened and had a relative decreased risk of death from breast cancer of about 16% (RR 0.84, 95% CI 0.74-0.95).

Because about 20% of the women in the AGE study did not actually obtain mammograms and 10% missed screening, the potential for risk reduction for women strictly compliant was probably an underestimate, they said.

Addressing the overall issue of early screening, Drs. Djulbegovic and Lyman noted that false negatives might lead to inappropriate reassurance and delays in diagnosis, whereas false positives could lead to unnecessary biopsies and additional imaging.

However, the editorial writers were especially concerned about the potential for radiation risk. While the overall reduction in death from breast cancer during the 10 to 15 years of follow-up in trials of screening mammography is clear, the anticipated peak for radiation-induced breast cancer occurs some 10 to 20 years after exposure. Furthermore, the risk might persist throughout a woman's life, they said.

The authors concluded that although the best estimates suggest that the harm seems to be less than the benefit, the risk of harm remains too uncertain to conclude with confidence that mammography in this age group has a consistent net benefit.

"Every woman, with her physician's guidance, should decide whether regret will be greater if she develops breast cancer that could have been detected earlier by screening mammography, or if she develops breast cancer later in life as a result of screening mammography itself," they concluded.

In a response to the findings of the AGE study, Robert Smith, Ph.D., director of cancer screening for the American Cancer Society, critiqued the findings.

There are a number of problems with this study that likely explain the results at this time, he said.

• The power of the trial was significantly diminished by attrition and nonattendance and by a lower than expected mortality in the control group.
• Decreased mortality is more likely due to increased awareness about the importance of breast cancer detection in the United Kingdom beginning in the late 1980s.
• The screening protocol used in the Age study was essentially a 1980s protocol, which has lower sensitivity to detect breast cancer lower than modern mammography.
• After the first screen, women were offered only single-view mammograms, diminishing sensitivity, especially in younger women with greater breast density. Today in the United Kingdom two screens are used in all mammography.

Turning to the issue or radiation exposure, Dr. Smith said the risks are theoretical and based on extrapolations from individuals exposed to large doses, either via atomic weapons or prolonged radiation exposure. Most commentaries about radiation risk, however theoretical the estimate, generally conclude that the risk-benefit equation is significantly in favor of mammography.

The authors also noted, he said, that their findings are consistent with other trials, but most of these results seem to be more similar to studies done in the 1970s. Second-generation trials in Sweden, for example, following more modern screening protocols, show benefits ranging from 36% to 44% mortality reductions.

"While we look forward to continued follow-up of these data, the results of this study are readily explained and do not substantively call into question the value of mammography for women in their forties," Dr. Smith said.

Current evidence supporting mammograms is even stronger than in the past, he said, adding, "Women can feel confident about the benefits associated with regular mammograms for finding cancer early."