Voir en Français

Small benefits, substantial harms with mammography screening

It’s time to accept the evidence

A version of this commentary appeared in the National Post and Huffington Post

We hear it often: “Early detection saves lives”— a promise that screening reduces breast cancer deaths.

While this statement is factually true, nevertheless, for every 10 women destined to die of breast cancer without screening, 7 or 8 of these women will still die even with screening.  This is what a breast cancer mortality reduction of 20 to 30% means.

After appraising data accumulated over more than 30 years, the Canadian Task Force calculated that  2100 women age 40-49 must be screened every two years for 11 years, so that one woman will avoid death from breast cancer. This estimate is consistent with guidelines from the US Preventive Services Task Force in 2009.

The miniscule benefit (0.05%) must be balanced against the downside of screening this age group (women aged 40-49). Fully 700 women (of the 2099 who had no benefit from screening) will have false-positive mammograms resulting in needless anxiety, unnecessary imaging and painful and intrusive biopsies after which they will be told they don’t have breast cancer.

Much worse will be the 10-15 women who receive a false diagnosis of breast cancer (over-diagnosis) and these women will experience the misery of unnecessary breast cancer therapy: surgery, radiation and chemotherapy, plus the burden of life-long concern about a cancer they never really had (see article).

The good news is that, according to growing evidence: we don’t need routine mammography to save even that one life.

Breast cancer deaths have been declining in countries which don’t screen routinely.  Numerous studies have shown that as early as 1985-1990 breast cancer death rates were declining in countries even before they offered screening programs.

Furthermore, breast cancer mortality rates have declined to an equal extent in populations with and without screening programs.  In the UK and Europe where screening is offered only to women 50 and over, mortality rates have declined in women age 40-49 by up to 37%.

Most remarkably, a decline in breast cancer mortality has been observed in women age 30-39, women too young to be screened in any screening programs.

In spite of this evidence from many western countries, including the US, screening advocates persist in attributing observed mortality declines to screening, not to improved therapy.  In fact, the evidence is strong that it is therapy, not screening, that accounts for the mortality decreases.

Another harsh truth is that screening has not reduced the incidence rate of advanced cancers in developed countries, although that was the expected consequence of screening (CMAJ Nov 22, 2011;183:1957-1958.)

Unexpectedly, screening has increased the overall incidence of breast cancer, both invasive and ductal carcinoma in situ.  This is largely attributable to over-diagnosis, estimated to be 20 to 50% of all screen-detected breast cancers by researchers outside the imaging industry.

Screening invitations promise that early detection means less radical surgery will occur because the tumours are smaller.  Yet a woman’s risk of getting a mastectomy (the removal of the whole breast) for a screen-detected cancer increases by up to 20% compared to women in similar populations where screening is not offered.

In a country where screening programs have been initiated, as in the UK, screening advocates like to point out that the mastectomy rate is lower in screened women than in those who refused to be screened.  Such a comparison is not good science, because those who go for screening tend to be more health conscious, more prosperous and more empowered displaying the healthy volunteer effect.  Non-participants do not represent a comparable population.

Advocates for screening also claim that early detection achieves increased survival time.  Early detection may result only in more years of living with disease compared to a woman detected later clinically.  That is, if both women end up dying of breast cancer, no additional years were gained by early detection.

The important outcome when assessing screening benefits should be reductions in deaths from breast cancer, or even better, a reduction in mortality from all causes.  The latter has never been shown.  Furthermore, when women are over-diagnosed with breast cancer, they will falsely be regarded as long-term survivors.

Finally, and in the midst of this compelling and complex data, we cannot overlook the issue of conflicts of interest.  For decades, the imaging industry has publicly excoriated panels who dared to evaluate the effectiveness of early screening, such as the NIH Consensus Conferences , the Cochrane Collaboration, the Canadian Task Force on Preventive Health Care and the United States Preventive Services Task Force — all indisputably objective, multi-disciplinary and expert.  Their guidelines were antithetical to the industry’s interests.

It is no wonder the public is confused.

The truth of more than four decades worth of research is now very clear: the potential benefit of mammography screening is small and the harms are substantial at all ages, but especially so for women in their 40s.

The bottom line is that mammography screening, implemented to reduce breast cancer deaths due to earlier detection of breast cancer, has been eclipsed by therapy and increased awareness.

Cornelia Baines is a Professor Emerita at the Dalla Lana School of Public Health, University of Toronto, and an advisor to EvidenceNetwork.ca, a comprehensive and non-partisan online resource designed to help journalists covering health policy issues in Canada. 

January 2012

 

Print Friendly, PDF & Email

This work is licensed under a Creative Commons Attribution 4.0 International License.