New Book: Mammography Screening—truth, lies and controversy

What happens when a popular cancer screening technology is found to be far more harmful than lifesaving? When the finding becomes clear decades after it was oversold to the public? When a lucrative industry, in terms of equipment, breast biopsies, drugs, etc., has already built around it that is now impossible to dismantle?

One might hope that science would win out. After all, mammography has the distinction of being a cancer screening test with extensive research behind it. In his new book Mammography Screening: Truth, Lies and Controversy (Radcliffe Publishing, London/New York: 2012), physician and research scientist, Peter C. Gøtzsche recounts what it was like to take a hard look at that research and find it didn’t match up with mammography’s sterling reputation.

The near-universal reaction? Shoot the messenger. Vicious attacks came from researchers, policymakers, and physicians. Too often aimed at the man himself rather than his critique. Opinions were fixed—mammography is risk-free and lifesaving. Anyone who disagrees publicly is causing deaths in women who might reconsider and stop having mammograms. The book describes the scientist’s 11-year investigation that uncovered mammography’s considerable harms, though they were “hiding” in plain sight—in the original studies that had long ago established mammography screening as a lifesaver.

Dr. Gøtzsche, director of The Nordic Cochrane Centre, Copenhagen, describes himself as someone who knew little about mammography when, in 1999, he was asked by the Danish Research Council to do an in-depth assessment of all mammography-related research. A statistician and expert in clinical trial design and analysis, Dr. Gøtzsche was the right man for the job. Denmark was considering a national screening program, but first wanted to know more. Bad signs were already showing up in Norway where such a program was underway. Screening decreased breast cancer deaths but, ominously, it hadn’t decreased the rate of deaths from all causes. Even more alarming, mammography failed to detect the most aggressive, deadly form of breast cancer.

Central to Dr. Gøtzsche’s conclusions are the nine randomized clinical trials that included a half million women altogether. The first took place in New York City, in the early 1960s; the last two trials were conducted in Canada and Sweden in the 1980s. “We were baffled by what we found,” he wrote. “We had expected them to be more convincing considering how popular mammography screening had become, despite its high cost.”

The results of these nine trials focused narrowly on mammography screening’s role in reducing breast cancer deaths. Dr. Gøtzsche may well be the first to step back and look at the big research picture, assessing the total death rate and the harm to women. His assessment for the Danish National Board of Health described the benefits as uncertain and raised the possibility that screening could cause more harm than good. It was ignored.

Dr. Gøtzsche continued mining the data from the nine trials and publishing frequently over the next decade. The first paper, co-authored with statistician Ole Olsen, appeared in 2000 in the British journal, The Lancet. But it was their second paper for The Lancet in 2001 that set off a furious international reaction. The nine mammography trials emphasize the number of breast cancer deaths among the participants, but Olsen and Gøtzsche contend that deaths from other causes must also be taken into consideration. These trials show that many more women given regular mammograms are treated for breast cancer than the unscreened women, and these treatments themselves may cause fatalities. Furthermore, overtreatment of ductal carcinoma in situ, often with mastectomy, was identified as “a considerable risk of mammography screening because most cases do not become invasive.” (Disclosure: I serve on The Nordic Cochrane Centre’s advisory board, am quoted in this book, and have reported Dr. Gøtzsche’s work ever since I first came across it in 2000.)

Reactions in the U.S. media were exceptionally virulent and prolonged. It was likely the first time that most physicians as well as the general public heard that some cancers will never cause death or symptoms. But this was not the first high-decibel mammography media controversy. In 1992, when the Canadian trial was published, it was roundly trashed because it came up with an unpopular finding: Mammography screening did not reduce breast cancer deaths, though it increased the number of cancers detected. Dr. Cornelia Baines, co-director of this trial, expected fellow scientists to take a dispassionate look at the finding to see why it differed from that of the earlier trials.  Instead, she became the target of numerous attempts to silence and discredit her.

When the mammography controversy surfaced again in the media in 2001, it was the policymakers, the radiologists, and the breast cancer specialists who came down hardest on Olsen and Gotzsche. To accept their conclusions would mean that hundreds of thousands of women worldwide have been treated for a type of breast cancer that would either regress or remain dormant. Who would “dig deep” into that possibility? Certainly not the doctors who for years have been sending their patients for mammograms. And certainly not the radiologists whose income had increased mightily—less from the screening test itself than from the money-making ancillary activities like stereotactic needle biopsies, continuing education courses, magnetic resonance imaging, and biopsy-related patents (click here for one example).

Most women don’t want to hear about mammography’s harms either. Fear of breast cancer sold them on mammography in the first place—without it, there would be no action to take. In the early 1970s when mammography screening was first introduced in the U.S., most American women were not particularly fearful of breast cancer, largely because it was seen as an old woman’s disease. But a multi-national cancer drug maker took care of that “problem” with annual breast cancer awareness campaigns featuring young breast cancer victims. The fear level is kept high for doctors, too, who are frequently reminded that “failure to diagnose breast cancer” is a leading cause of malpractice lawsuits.

Cancer charities take a well-deserved hit in this book for their refusal to admit that screening has a downside. Their misuse of statistics seems calculated to inflate the benefit of cancer screening. Consider the 30% reduction in deaths bandied about in the early years of mammography promotion. This statistic was downgraded recently to 15% by the U.S. Preventive Services Task Force. But both of these are relative risk statistics, which are typically misunderstood by doctors and consumers alike. Most relevant is the absolute effect of screening, not the relative effect, points out Gøtzsche who provides this explanation: “If 2,000 women are screened regularly for 10 years, 1 woman will avoid dying from breast cancer, and 10 healthy women who would not have been diagnosed without screening, will have breast cancer  diagnosed and be treated unnecessarily.”

At the end of last year, the Canadian Medical Association Journal invited Dr. Gøtzsche to write an editorial entitled, “Time to stop mammography screening?”  The Canandian Task Force on Preventive Health Care had just issued new guidelines,  stating that  “women who do not place a high value on a small reduction in breast cancer mortality, and who are concerned with false-positive results on mammography and overdiagnosis, may decline screening. ”  Dr. Gøtzsche describes this as “an important step in the right direction, away from the prevailing attitude that a woman who does not undergo screening is irresponsible.”

It’s hard to imagine that this could ever happen here in the U.S.

This book can serve as a guide to physicians and women who want to make their own informed decisions about mammograpy screening, who want an honest in-depth assessment of the research—one that should have given to the public before the introduction of mass screening. A similar “promote the test first, learn the harms later” story has unfolded recently about the PSA screening test for prostate cancer. You just might want to sharpen your critical skills and prepare in advance for the next cancer screening disaster.

Maryann Napoli, Center for Medical Consumers©

More about Dr. Gotzsche’s work:
Free mammography screening leaflet from the Nordic Cochrane Centre  It is also available  at The Nordic Cochrane Centre website in 13 languages.
Cut your risk of breast cancer—avoid screening mammograms. One-third of all breast cancers found on a mammogram are the forms of breast cancer that would never cause death or symptoms.
Breast cancer death rate has dropped, but not due to mammography  Improvements in breast cancer treatments are most likely cause. ‘Before and after’ studies conducted in countries that introduced mammography in the 1990s verify what was noticed in Norway in this era: Screening  does not detect the most deadly form of breast cancer; it has not reduced the occurrence of advanced cancers.
Poster for the 2002 Cochrane Colloquium  U.S. media coverage of the 2001 Lancet paper.

Breast cancer deaths drop—but not because of mammography

Mammography screening is usually credited with the drop in breast cancer deaths recorded in many countries, including the U.S.  But a case is building for improvements in breast cancer treatment as the most likely cause. The decrease in deaths has occurred in many European countries that did not start  mammography screening until the 1990s, which happens to coincide with greater use of long-term adjuvant therapy (e.g., tamoxifen, chemotherapy) given after the initial treatment is over.  Researchers say the case for adjuvant therapy is made stronger by the fact that,  in some of these countries, the greatest decreases in breast cancer deaths were among young women (under 50), the age group that never received mammography screening.

As someone who has followed the “selling” of mammography screening to American women that started in the early 1970s, I offer some background for the new findings from Europe. Thanks to nearly 50 years of research, we know more about mammography than any other cancer screening test. Expert panels with no conflict of interest have concluded that the breast cancer mortality rate among mammography-screened women (in randomized trials) is only 16% lower than that of unscreened women. In the U.S., there was no reduction in breast cancer deaths until the early 1990s and about 2% a year thereafter.

Today, there is a greater understanding of cancer. Some abnormalities that look like cancer under the microscope do not become invasive, if left untreated. Many regress spontaneously, stay put, or grow so slowly they will never make their presence known.  At least as far back as the 1970s, pathologists knew about these non-progressive cancers that can occur in all major organs of the body.  But women weren’t hearing from them.  Instead, radiologists and surgeons dominated the promotion of mammography screening in the early years. Today, it is the radiologists who are often quoted in the media, warning us about the dangers of forgoing mammography screening while downplaying its harms.

Well, it is quite reasonable for women to forgo screening—that is, after becoming well-informed. Here are the highlights of several review articles published in the last few weeks.  See below for my sources.

• There are usually dramatic increases in the discoveries of new breast cancers after mammography screening takes off. Tomorrow’s cancers are found today is the standard explanation.  This  ignores the fact that in every major randomized trial, some of the regularly screened women—who have had many previous “all-clear” mammograms—are nonetheless diagnosed with  invasive tumors that are fatal despite prompt treatment.  Recent studies conducted in many countries, including the U.S., show that mammography screening has not reduced the occurrence of large invasive cancers.

• The aforementioned large increase in new cases of breast cancer without a large decrease in the rate of new cases of advanced cancer (over time) indicates that much of the increase is due to detection of non-progressive cancers (i.e., overdiagnosis). Here’s how the Cochrane review on mammography screening assessed the damage: “For every 2,000 women who are screened throughout 10 years, one will have her life prolonged. In addition, 10 healthy women, who would not have been diagnosed if there had not been screening, will be diagnosed as breast cancer patients and will be treated unnecessarily.”

• Women are often told that mammography-detected breast cancers require less drastic treatment.  The opposite is true. Mastectomy rates were going down in some European countries in the years prior to the introduction of mammography screening but went up afterwards. Many countries, including the U.S., show 20% more mastectomies in the screened women compared to the unscreened women. One factor is the large increase in the detection of ductal carcinoma in situ, or stage 0 breast cancer, which was rare in all countries prior to the introduction of mammography but is now a common diagnosis.  DCIS  is treated increasingly with mastectomy, though it has long been known that only about 20% to 30% of DCIS will go on to become invasive breast cancer if left undetected, according to Susan Love, MD, author of Dr. Susan Love’s Breast Book.

• Physicians at the Dartmouth Institute of Health Policy addressed a common misconception about mammography in a paper published online in Archives of Internal Medicine (see below). “The presumption often is that anyone who has had cancer detected has survived because of the test, but that’s not true,” according to co-author H. Gilbert Welch. “In fact, and I hate to have to say this, in screen-detected breast and prostate cancers, survivors are more likely to have been overdiagnosed than actually helped by the test …  It’s important to remember that of the 138,000 women found to have breast cancer each year as a result of mammography screening, 120,000 to 134,000 are not helped by the test.”

Maryann Napoli, Center for Medical Consumers©

Related post
There is aid for breast cancer patients who must decide about adjuvant therapy.  read more

Sources for above post
“Why mammography screening has not lived up to expectations from randomized trials.” Cancer Causes Control, published online November 10, 2011.  This is the source for virtually of the above.  Click here

Cochrane review that assessed the harms as well as benefit of mammography screening.  Click here   Want to know more about the Cochrane Collaboration?   Click here

“Likelihood that a woman with screen-detected breast cancer had her ‘life saved’ by that screening” Archives of Internal Medicine, published online October 24, 2011.  This is the source for the last bullet point.  click here    Better yet, Click here for an easier to read New York Times article about this study.  Added December 4, 2012:  YouTube vide0 explaining a new study that found mammograpy screening accounts for overdiagnosis and overtreatment of 1.3 million American women over the four decades since it was first introduced.

Breast Cancer Awareness Month: Read This Before You Have a Mammogram

Mammography Screening—Both Good and Bad News

Mammography screening seems destined to spawn controversial study results, as it has in the last few months. And its role in reducing the breast cancer death rate will remain the subject of heated debate among researchers for a good long time. The U.S. breast cancer death rate has declined 20% since 1990, as it has in other western countries. But it is not clear whether this can be credited to mammography, treatment improvements, or simply greater awareness among women that breast tumors should be promptly examined. What has become increasingly clear is this: Mammography screening harms more women than it helps.

That’s not the message that goes out to women via the media and their doctors. Of the three recently published mammography screening studies, the most media attention went to the one that seemed to equate the improvements in breast cancer survival with the increase in the proportion of women now diagnosed with small breast cancers. That was the take-home message from a CNN TV interview with Clifford Hudis, MD, a co-author of this study published last month in the journal Cancer.

With his colleagues at Memorial Sloan-Kettering Cancer Center, New York City, Dr. Hudis reviewed the early-stage breast cancers diagnosed in the U.S. over a 24-year period, from 1975 to 1999. The statistics came from the government’s database that draws from nine cancer registries across the country considered to be broadly representative of the U.S. population.  Information concerning more than ¼ million breast tumors from the same population was also available to the researchers. The upshot: Many more small breast cancers were diagnosed in 1999 than were diagnosed in 1975.

Certainly sounds like good news, and that’s how it played out in the Wall Street Journal and other media. Yet despite the messages conveyed in the CNN TV interview, Dr. Hudis’s study did not determine why the breast cancer death rate went down. It simply found a greater number of women diagnosed with breast cancer in recent years had small tumors.

It is unclear why Dr. Hudis would misrepresent his own study and leave CNN viewers with the impression that mammography and breast self-examination (BSE) are largely responsible for the decline in breast cancer deaths. Let’s put aside for the moment Dr. Hudis’s discredited belief in BSE, as two trials showed the practice leads to unnecessary biopsies with no discernible benefit.

Instead, a hard look at the nation’s increase in small cancers is long overdue. The topic is too complicated for a four-minute TV interview. The breast cancer death rate may be going down, but there has also been a steady annual increase in the number of women diagnosed with breast cancer over the last two decades. The increase is generally acknowledged to be related, at least in part, to mammography screening. This is not entirely good news. Mammography screening may prolong life for some women, but it also causes many more women to be diagnosed and treated for cancers that they would be better off not knowing about. Not all cancers are deadly even if left undetected.

“Screening mammography is great for finding small cancers. But finding small cancers is not necessarily good,” according to Donald Berry, PhD, Chair of the Department of Biostatistics and Applied Mathematics at the University of Texas M.D. Anderson Cancer Center. In an e-mail interview, Dr. Berry explained that breast cancer comes in several different forms. “Some cancers are lethal and others are not. Non-lethal cancers tend to be smaller, and they tend to grow more slowly.” Dr. Berry recently co-authored a study that showed that breast cancers detected by mammography tend to have the best prognosis (excellent chance of recovery with treatment), even when the cancer has spread to the lymph nodes. The authors advise doctors to take that into consideration when recommending treatment.

Dr. Berry and colleagues assessed the outcomes of two major mammography screening trials that together included over 150,000 women, singling out those who eventually developed breast cancer. More cancers—both invasive and noninvasive (ductal carcinoma in situ)—were detected among mammography-screened participants, as compared with those not given mammograms.

“Screening mammography finds a greater proportion of non-lethal cancers than do other methods of detection. Unfortunately, it is not possible to perfectly identify which cancers are lethal. So we treat them all. Consequently, screening leads to overdiagnosis and excessive treatment,” Dr. Berry explained, referring to surgery, radiation, hormonal therapy, and chemotherapy that some women could have avoided without risking their health.  Unfortunately, women have been led to believe the opposite—that mammography leads to less drastic treatment.

One of the trials assessed by Dr. Berry and colleagues found the breast cancer death rate among the women given mammograms was no different from that of the women not given mammograms. This is after 16 years of followup. The finding clearly suggests that mammography is not only useless but harmful to a significant number of women. There were 82 additional breast cancers* (half were invasive and half were noninvasive) in the women given mammograms. In other words, 82 of the mammography-screened women were treated for a cancer, including invasive cancer, which would not have produced symptoms or become life-threatening.

No other cancer screening test has been as well studied as mammography. Of the seven such trials conducted worldwide, two found no difference in the breast cancer death rate between the mammography-screened women and those not given mammograms. In a systematic review of the seven mammography trials conducted in 2000 by the Cochrane Collaboration, the two trials that found no life-prolonging benefit to mammography were determined to be superior in terms of quality than the five trials that showed a modest benefit to mammography. (None were without flaws.) The Cochrane Review’s conclusion: “Screening is unjustified.”

The merits and flaws of these seven major trials have dominated the controversies over the value of mammography. Little is known about the quality of mammography screening in the real world. Women who participate in clinical trials are given high-quality mammograms that are read by experienced radiologists. In other words, it is mammography under the best of circumstances. Not enough research attention has been given to the question of how women fare when given mammograms in the everyday practice of medicine. That was the point of a new study led by Joanne Elmore, MD, School of Medicine, Harborview Medical Center, University of Washington, Seattle.

With colleagues at several medical centers, Dr. Elmore looked at the medical records of women enrolled in six health plans around the country. They singled out the women who died from breast cancer between 1983 and 1998, an era that followed major improvements in mammographic techniques. 1351 women with breast cancer (aged 40 to 65 years) were matched with 2501 cancer-free women who were the same age and with the same level of risk for breast cancer.

Dr. Elmore and colleagues compared the screening practices (mammography and a physician’s breast exam) of the women who died of breast cancer with the high-risk women who were still alive. High-risk was defined as having a family history of breast cancer or a breast biopsy, as noted in the medical records. If screening reduces the rate of breast cancer death, the researchers expected to find that the women who died of breast cancer had undergone less screening then the women still alive. In fact, there was no difference between the two groups.  “Our findings suggest that breast cancer screening in the community was minimally effective in preventing death,” concluded Dr. Elmore and colleagues. Results were published in the Journal of the National Cancer Institute.

In the editorial that accompanied this study, Russell Harris, MD, University of North Carolina, Chapel Hill, gave several reasons why screening may be making “a smaller contribution” to the reduction in breast cancer deaths than the major trials have led us to believe. “Better treatment may mean that screening is less necessary than it was previously, because treatment of later stage cancers may still be effective.”  He also cites the fact that women today are more likely to have breast lumps “found accidentally” promptly examined by a health professional. Here Dr. Harris is referring to women finding tumors while in the shower, as opposed to finding them while doing BSE.

Women continue to get one-sided information about mammography (“it will save your life”) and little about its harms (unnecessary treatment). This was reflected in a survey of American women’s attitudes toward this technology in 2000. Only 6% reported that they knew about cancers that do not progress or become life-threatening.

For an international perspective on the accuracy of the mammography information dispensed to women via Web sites in eight countries, a 2003 survey was conducted by Karsten Juhl Jorgensen, MD, and Peter C. Gotzsche, MD, of the Nordic Cochrane Center in Copenhagen. They assessed 27 Web sites that were sponsored by governments, professional organizations, and consumer advocacy groups.

Jorgensen and Gotzsche, whose findings appeared last year in the British medical journal, BMJ, concluded, “The information material provided by professional advocacy groups and governmental organizations is information poor and severely biased in favor of screening…and failed to mention major harms.”  On the other hand: “Web sites of consumer groups were more balanced and comprehensive.” The three consumer advocacy groups with the balanced information are all U.S.-based: The National Breast Cancer Coalition (www.stopbreastcancer.com), Breast Cancer Action (www.bcaction.org) and the Center for Medical Consumers (www.medicalconsumers.org).

Based on the findings from seven clinical trials, Jorgensen and Gotzsche describe the cost to women in terms of unnecessary treatment. For every 1000 women who undergo mammography screening for ten years, they report, one woman will have her life prolonged; five additional women will receive an unnecessary cancer diagnosis and treatment; and three women will have a benign tumor biopsied.

“Whether this is too high a price to pay is open to debate,” wrote Jorgensen and Gotzsche, “but if women and policy makers are not informed of this balance between major benefits and major harms—which they have not been so far—then there cannot be any discussion or rational decision making.”

Bottom Line: Mammography-detected breast cancers have the best outlook. The screening test also leads to the detection and treatment of breast cancers that would never become life-threatening. Mammography’s role in the nation’s declining breast cancer death rate remains unclear. At best, it appears minimal. Women are not receiving honest information about mammography’s harms.

*There were 592 cases of invasive breast cancer and 71 noninvasive breast cancer in the approximately 25,000 women given mammograms, as compared with 552 invasive and 29 noninvasive breast cancers in the approximately 25,000 not given mammograms. The women were age 40 at the start of the trial. Only half of all noninvasive cancers, also called ductal carcinoma in situ, are likely to become invasive, but all are treated as if they will. This trial indicates that invasive cancers do not always progress and become lethal.

Maryann Napoli, Center for Medical Consumers ©