Rethinking the screening mammogram

Robert H. Shmerling, MD

Faculty Editor, Harvard Health Publishing

Doctors and patients want and need tests that detect disease while that disease is still treatable — ideally, while curable. Even better, we want screening tests that can accurately detect “pre-disease,” not just detect it once present. An example is colonoscopy, which can detect certain types of polyps that degenerate into colon cancer. And of course, we want tests that actually improve longevity or quality of life.

On the other side, we don’t want too many false alarms (or “false positive” results) — these are results that suggest disease when, in fact, no disease is present. This is especially true if those false alarms lead to unneeded surgery or other risky procedures.

Even when risky treatments are avoided, there’s another downside to false alarms: worry. If you’re told your screening test for cancer is abnormal, the time it takes to figure out that it’s a false alarm can be terribly frightening. And it can seem like forever.

Assessing the value of the screening mammogram

In a previous blog, I wrote about how we have relatively few time-tested and reliable medical screening tests that deliver on their promise. A 2016 study questions the value of one of those tests: the mammogram.

This study analyzed data from women over 40 and compared the size of breast cancers at the time of diagnosis detected in the 1970s (before mammography became common) with the size of tumors detected between 2000 and 2002, when screening mammography was routine. Treatments and rates of death due to breast cancer 10 years after the diagnosis were also analyzed. The study found that:

  • As more women underwent routine screening mammograms, more small breast cancers were detected. Many of these tumors were restricted to the ducts within the breast (called ductal carcinoma in situ), and even without treatment would never threaten the health of the woman.
  • The detection of larger, more aggressive breast cancers was unchanged in frequency between the pre-mammogram and more recent time periods. This is important since, if screening mammograms caught these cancers earlier, in theory, the frequency of detecting these more dangerous tumors should be falling.

Are there benefits of a screening mammogram?

The study’s authors did not suggest that mammography is useless. They estimated that about 20% of women with small tumors that could only be detected with a mammogram received therapy that might be lifesaving. But the other 80% of women did not benefit. Similarly, estimates suggest that about two-thirds of the reduction in breast cancer deaths in recent years is due to better treatments, not better detection. Some additional reduction may be related to falling rates of post-menopausal hormone replacement; but the contribution of mammography to this trend may be relatively small.

The authors suggest that the benefits of mammography have been overestimated and, as a result, many women may be receiving medical and surgical treatments that are unnecessary. It’s not surprising that this new study is controversial. A number of experts have criticized its conclusions and worry that such research will discourage women from having mammograms that could save their lives.

Debate on the value of screening mammograms has happened before

This isn’t the first time questions have been raised about the usefulness of mammography. A 2012 study also raised the concern that unimportant tumors were being detected, leading to unnecessary treatment for large numbers of women. There has also been controversy about when women should start having them and how often they should be repeated. For example, the US Preventive Services Task Force suggests that routine screening mammograms should begin at age 50 and be repeated every two years until age 74. But the American Cancer Society says that women should begin at age 45, have them yearly until age 55, and then every two years.

So, what’s a woman to do?

While the message of this new study is clear — mammography may not be as good as we thought — it isn’t at all clear what women should do with this information. It’s likely that in light of these concerns, recommendations will change in the future. But how?

Optimal recommendations regarding mammography will depend on:

  • Developing better ways to distinguish breast cancers that pose a threat to health and warrant aggressive therapy from those that don’t
  • Understanding the individual factors that affect the usefulness of mammography. For example, family history, genetic factors, use of hormonal therapy, breast density, obesity, and reproductive history all can affect the risk of breast cancer or the usefulness of mammography.
  • Reaching a consensus regarding an acceptable level of accuracy for screening tests. As screening tests for breast cancer evolve — including better mammograms, ultrasound, and MRI — the accuracy of diagnosis should improve. But since no test is perfect, we have to accept that some breast cancers will be missed and false alarms will happen, even with the best available screening methods.

For now, it’s important to talk to your doctor about this important study. My guess is that he or she will continue to recommend mammograms as before.

Still, it’s worth a conversation. There is a real possibility that screening tests such as mammography can cause harm — and that’s worth understanding before you have the test.

Follow me on Twitter @RobShmerling

Comments:

  1. Daniel Kopans, M.D.

    I am a Professor of Radiology at Harvard and one of the leading experts in breast cancer screening.

    I would urge Dr. Shmerling to reread the paper by Welch et al to which he refers in this Blog that led him to cast doubt on the value of mammography screening. As the authors point out in that paper, their argument is based on the false claim that the “underlying disease burden was stable” over the period that they reviewed. A careful review would reveal that Dr. Welch is not hampered by facts. He used the exact same data in his 2012 paper in the same journal (Bleyer A, Welch HG. Effect of three decades of screening mammography on breast-cancer incidence. N Engl J Med. 2012 Nov 22;367(21):1998-2005) to conclude that the underlying disease burden had been increasing by 0.25% per year to 0.5% per year. He was, in fact, incorrect in that paper as well as this more recent paper. The fact is that the incidence of invasive breast cancer (based on the Connecticut Tumor Registry that predated SEER) had been increasing by 1-1.3% per year for decades (starting in the 1940’s). Numerous studies have accepted this fact, but since it does not suit his biases, Dr. Welch has ignored the data. He claims to rely on SEER data, but keeps ignoring the fact that SEER began in 1974 in the same year that the wives of the President and Vice President of the U.S. were both diagnosed with breast cancer. Many women rushed to get mammograms and this is reflected in the small prevalence peak in 1974. This ad hoc screening ended quickly and would generate a compensatory decline in breast cancer incidence (cancers that would have become evident in 1975, 1976, and 1977 had already been found by screening). Consequently, the period that Dr. Welch likes to use for his “baselines” is the most unreliable in the SEER data.

    The underlying disease burden is the crux of the arguments against screening. If the underlying rate of invasive cancers has continued to increase (there is no reason to expect that it suddenly stopped after consistent increases since 1940) then Welch and his arguments of massive overdiagnosis are false, and his claim that there has been little reduction in advanced cancers is false. No one has ever seen an invasive breast cancer found by mammography disappear on its own, yet Welch claimed there were 70,000 in 2008 alone! 70,000 in one year and no one has ever seen it happen!!?

    Mammography screening is not the ultimate answer to breast cancer. It does not find all cancers and does not find all cancers at a time when cure is possible, but it is the main reason that deaths have declined dramatically since 1990. Advanced breast cancer is incurable. The only way to reduce deaths is to treat breast cancer early and this is why screening is so important.

  2. Robert Shmerling

    MRIs do not expose people to radiation.

    Doctors have been encouraging patients to stop smoking for decades – is there more doctors should be doing in this regard?

    Prevention of breast cancer should be a major priority – unfortunately, ways to reliably prevent most breast cancers have not been discovered; maintaining a healthy weight, choosing a healthy diet, moderating alcohol intake, not smoking, avoiding excessive estrogen exposure may help but these are already recommended.

    I think most doctors would strongly recommend avoiding/limiting exposure to carcinogenic toxins. The controversy comes in figuring out how to do that.

  3. SusanB

    Yes, after reading this article and others, it isn’t at all clear what a woman should do. I have had 7 or 8 mammograms in my life, starting in my late 30’s as I had a very “thorough” PCP at the time. Now at 57, I am scheduled for a 3D mammogram next month. I am not sure I wish to go through with it as I fear the advanced technology will lead to something atypical or DCIS, MD’s are not sure “what it means” but will recommend aggressive treatment “just in case”.

    And will the aggressive treatment, e.g. radiation, tamoxifen, etc. lead to possible further medical problems in the future?

    My MD has said “you can always get the biopsy”. That is also not reassuring based on what I’ve read.

    Can I just ask for an ultrasound and skip the mammogram? This is also something that is not clear. Guess I have to do some more research.

  4. Robert H. Shmerling, MD

    MRIs do not expose people to radiation. (Xrays and CT scans do, however)

    Doctors have been encouraging patients to stop smoking for decades – is there more doctors should be doing in this regard?

    Prevention of breast cancer should be a major priority – unfortunately, ways to reliably prevent most breast cancers have not been discovered; maintaining a healthy weight, choosing a healthy diet, moderating alcohol intake, exercising, not smoking, & avoiding excessive estrogen exposure may help but these are already recommended.

    I think most doctors would strongly recommend avoiding/limiting exposure to carcinogenic toxins. The controversy comes in figuring out how to do that.

  5. azure

    MRIs have health effects as well. More exposure to various levels of radiation isn’t helpful unless there’s a very good reason for it, so why push MRIs?

    When will MDs and blogs such as this discuss the issue of prevention as other then not smoking tobacco? How about MDs pushing for avoidance of endocrine-disruptors like pthalates? Or decreasing exposure to pesticides and other toxic, mutagenic, carcinogenic, teratogenic compounds?

    Too controversial?

Post a Comment:

This blog aims to provide reliable information as well as healthy dialog about the topics covered. We reserve the right to delete comments for any reason, particularly those that do not relate directly to the contents of this post, are commercial in nature, contain objectionable or inappropriate material, or otherwise violate our Privacy Policy. Promotional URLs will be removed from comments. Comments on this blog do not represent the views of our editors or Harvard University, and have not been checked for accuracy. All comments submitted to this site become the non-exclusive property of Harvard University.