Every year in the United States, hundreds of thousands of women are diagnosed with an early stage of breast cancer. Almost all believe that this early diagnosis has saved their lives. As it turns out, only one in four is right. The other three have to suffer the complications of surgery, radiation therapy, hormone therapy, and chemotherapy with no benefit. Most of these early breast cancers are actually precancerous tumors that may or may not progress to true cancers. How can we tell who benefits from these early diagnoses and who doesn’t?
Breast cancer is the most common cancer in women worldwide. Every year, more than 1.5 million women are diagnosed with breast cancer and 500,000 die from the disease. Breast cancer is second only to lung cancer as the most common cause of cancer death in women.
In the United States, more than 200,000 women are diagnosed with breast cancer every year, and 40,000 die from the disease. Indeed, one in eight women will be diagnosed with breast cancer. When aggressive, the cancer spreads to the liver, brain, bone, and lungs. For this reason, it’s important to diagnose it early.1
In 1895, Wilhelm Röntgen detected and produced X-rays, electromagnetic radiation of high energy and short wavelengths that pass through anything opaque to light, including human skin. More than fifty years later, Robert Egan, while working at the University of Texas MD Anderson Cancer Center, modified Röntgen’s discovery. Using much lower levels of radiation, Egan was able to detect abnormal lumps as small as a fraction of an inch, tumors that could never have been detected by physical examination.2 In 1958, Egan published a paper about it, and in 1964 he published a book, Mammography. The name stuck.
The first person to champion mammography was a radiologist named Philip Strax. In the early 1960s, in collaboration with Sam Shapiro, a statistician, and Louis Venet, a surgeon, Strax performed a study proving that mammography could detect breast cancers undetectable by physical exam. Mammography, argued Strax, could save lives. Then he proved it.
In December 1963, working with the Health Insurance Plan of Greater New York, Strax, Shapiro, and Venet divided 62,000 women between ages forty and sixty-four into two groups. One group received yearly mammograms; the other didn’t. Six years later, the researchers found that thirty-one women in the mammography group had died of breast cancer and fifty-two in the no-mammography group. In a paper published in the Journal of the American Medical Association, the researchers concluded, “The present communication offers the first hard evidence on the impact that screening has on mortality and the results are encouraging.”3
This was a breakthrough study. Between 1935 and 1967, the death rates for women from all causes had decreased by 42 percent. The death rate from breast cancer, however, hadn’t budged. The Strax study offered the first ray of hope that things could change. In 1988, Strax and Shapiro shared the prestigious Kettering Prize for outstanding contributions in the field of cancer.
More studies followed.
In 1985, researchers in Sweden divided 135,000 women age forty or older into two groups. One group was offered screening mammograms every two to three years; the other wasn’t. Subjects were followed for seven years. At the end of the study, those receiving mammograms who were age fifty or older had a 31 percent decrease in death from breast cancer.4
In 1993, researchers from medical schools in New Hampshire, North Carolina, and Maryland reviewed the results of eight studies that involved 500,000 participants. They, too, found that screening mammography had decreased the incidence of fatal breast cancer by about a third in those participants over age fifty. The authors enthused, “Randomized trials have provided stronger scientific evidence regarding the effectiveness of screening for breast cancer than for any other cancer” (emphasis mine).5
The results were clear. Mammography (a procedure that took only twenty minutes) could save lives. By the end of the 1980s, about 70 percent of U.S. women were getting them. The incidence of deaths from breast cancer steadily decreased.6
Then the picture darkened and blurred. As a consequence, some countries no longer recommend mammograms for their citizens. What happened?
In 2000, researchers from the Nordic Cochrane Centre in Denmark reviewed the results of eight studies of women who did or didn’t get mammograms, some of which were performed better than others. They concluded, “The two adequately randomized trials found no effect on breast cancer mortality or on total mortality.”7
In 2013, these same researchers reviewed the results of several more studies that followed 600,000 women ages thirty-nine to seventy-four who did or didn’t get mammograms and found that “The trials with adequate randomization did not find an effect of screening on total cancer mortality, including breast cancer, after ten years.”8
In 2014, Canadian researchers completed a twenty-five-year study in which they’d divided 90,000 women between ages forty and fifty-nine into two groups. One group received screening mammograms every year; the other didn’t. At the end of the study, 3,250 women in the mammography arm were diagnosed with breast cancer, and 500 died from the disease. In women who had never received mammograms, 3,133 were diagnosed with breast cancer, and 505 died from the disease—a statistically insignificant difference. And it didn’t matter which age groups were examined. Differences weren’t observed in women between ages forty and forty-nine or ages fifty and fifty-nine. No one, apparently, had benefited. The authors concluded that the “value of mammography screening should be reassessed.” Then they gave an important clue as to what was happening: “Annual mammography in women aged 40 to 59 does not reduce mortality from breast cancer beyond that of physical examination or usual care when adjuvant [additional] therapy for breast cancer is freely available” (emphasis mine).9 It was this last phrase that hinted at why researchers were now having so much trouble proving that mammography saved lives.
What did the authors of the 2014 Canadian study mean when they wrote “adjuvant therapy for breast cancer”? During the past twenty years, advances in the treatment of breast cancer have been remarkable, owing mainly to our ability to define various proteins on the surface of breast cancer cells. Now breast cancers can be divided into three types. Some breast cancer cells bear receptors on their surface for the hormones estrogen or progesterone. Others express a protein called human epidermal growth factor 2 (ERBB2, formerly Her2). Yet another group of cancers (so-called triple-negative breast cancer) doesn’t express any of these proteins. Different breast cancer types require different therapies. Today, in contrast to the time when mammography was first studied, cancer-targeted antibodies, hormone therapies, and chemotherapeutic drugs specific for different types of breast cancer are available. As a consequence, the five-year survival rates for early breast cancers that express either estrogen or progesterone or ERBB2 are 94 to 99 percent! For triple-negative cancers, the five-year survival rate is 85 percent.10 For this reason, even women whose breast cancers aren’t detected early have an excellent prognosis.
The impact of breast cancer treatments becomes clearer when examining mortality rates in different parts of the world. In resource-rich countries, such as the United States, the five-year survival rate for breast cancer exceeds 80 percent. In resource-poor countries, where therapies are less available, the five-year survival rates are between 10 and 40 percent.11
At this point, one could argue that at worst, mammography provides a marginal benefit in preventing deaths from breast cancer. But there’s a bigger problem. Early screening with mammograms has caused many women to suffer needlessly. No study showed just how damaging mammography can be more than one published in the New England Journal of Medicine several years ago.
In 2012, Archie Bleyer, from the Oregon Health and Science University, in Portland, and Gilbert Welch, from Dartmouth’s Medical School, published a paper titled “Effect of Three Decades of Screening Mammography on Breast Cancer Incidence.”12 Bleyer and Welch found that since the inception of breast cancer screening programs in the United States, the number of cases of early breast cancer had doubled, from 112 to 234 cases per 100,000 women screened. Therefore, every year, 122 more women per 100,000 were being diagnosed with breast cancer. On the surface, these numbers were encouraging. Now more than 100 women per 100,000 no longer had to die from breast cancer. But that’s not what these researchers found. Rather, they found that the rate of late-stage breast cancer had decreased by only 8 percent, from 102 to 94 cases per 100,000. This meant that only 8 of those 122 women who had been diagnosed with early-stage breast cancer had been prevented from progressing to late-stage disease. In other words, only 8 women per 100,000 had had their lives saved by early screening. But what about the other 114 women? Why hadn’t they had their lives saved by mammography? One possibility is that these early cancers had either receded on their own or were so slow growing that the women would have invariably died from something else; meaning they died with their breast cancer, not from it. Another possibility is that the cancer was so easily cured that it didn’t matter when it was detected—a consequence of better treatments. A final possibility is that the cancer was so virulent that it didn’t matter when it was detected; these women would have died from their breast cancer independent of whether the diagnosis had been made early or late.
One thing, however, was clear from Bleyer and Welch’s study. At least 114 women per 100,000 had undergone a combination of surgery, radiation, chemotherapy, or hormonal therapy without benefit. Bleyer and Welch estimated that during the previous thirty years, about 1.3 million women in the United States had been treated aggressively for breast cancers that never would have killed them, and that about 70,000 women were continuing to be overdiagnosed and overtreated every year. The authors concluded that “Despite substantial increases in the number of cases of early stage breast cancer detected, screening mammography has only marginally reduced the rate at which women present [to their doctor] with advanced breast cancer. Although it is not certain which women have been affected, the imbalance suggests that there is substantial overdiagnosis, accounting for nearly a third of all newly diagnosed breast cancers, and that screening is having, at best, only a small effect on the rate of death from breast cancer” (emphasis mine).
Fewer women today are dying from breast cancer. The Bleyer and Welch study had shown why. “We are left to conclude, as others have, that the good news in breast cancer—decreasing mortality—must largely be the result of improved treatment, not screening,” they wrote. The bad news was that the physical and emotional cost of early detection with mammography was high. Surgeries are disfiguring. Radiation therapy increases the risk of heart disease and lung cancer.13 Hormone therapies cause a variety of unpleasant side effects. And chemotherapies cause anemia, nausea, vomiting, hair loss, bone thinning, heart toxicity, and immune dysfunction leading to severe and occasionally fatal infections.
Bleyer and Welch’s findings were reproduced by others. In 2012, researchers from Harvard Medical School evaluated a nationwide screening program in Norwegian women ages fifty to sixty-nine. They found that for every 2,500 women screened over ten years, 1 death from breast cancer had been prevented. Unfortunately, 6 to 10 women were overdiagnosed and overtreated, all no doubt wrongly believing that their lives had been saved by aggressive therapies. The authors concluded, “Overdiagnosis and unnecessary treatment of nonfatal cancer creates a substantial ethical and clinical dilemma and may cast doubt on whether mammography screening programs should exist.”14
In 2015, researchers from the University of Washington School of Medicine found that for every 10,000 women screened over ten years, as few as 30 and as many as 194 would be overtreated for a cancer that never would have killed them. The authors worried that most women signing up for yearly or biennial mammograms wouldn’t know this, and they concluded that “physicians have been more likely to discuss benefits than harms of screening and frequently do not ask about women’s preferences.”15
In 2015, Archie Bleyer updated his estimates of overdiagnosis. Three years had passed since his New England Journal of Medicine paper was published. Now he estimated that every year, 78,000 women in the United States were being overdiagnosed and overtreated for a breast cancer that never would have killed them. He lamented that “most patient-education ads do not even mention overdiagnosis, and most women are unaware of its possibilities.” He also didn’t let doctors off the hook, arguing that by omission, women were being misinformed. “Radiologists, pathologists, and other professionals involved with screening mammography should recognize that the potential harm of overdiagnosis is downplayed or not discussed with the patient and family, despite agreement that the objective is informed choice.”16
No recent study has ever shown that mammograms have prolonged the lives of American women. This doesn’t mean that mammograms don’t save some lives. It only means that the number of lives saved by mammograms, given the advances in breast cancer treatments, hasn’t been enough to have had an impact on overall mortality.
Although overdiagnosis and the subsequent harms of unnecessary invasive treatments are the greatest problems with early screening, they’re not the only ones. Labeled a “false-positive screen,” this other problem refers to women who are asked to come back to the doctor due to suspicious findings on the mammogram only to find out later (sometimes much later) that they didn’t really have cancer.
In 1996, researchers from Stockholm, Sweden, looked back at a nationwide screening program that had begun in 1981. Thousands of women had been screened. The first round of screening was completed in 1983. The rate of false positives was astounding. The researchers found that 352 women had been diagnosed with a suspicious mammogram. As a consequence, these women made 1,112 repeat visits to the physician, suffered 397 initial biopsies or repeat biopsies, and underwent 187 confirmatory mammograms. After six months, 64 percent of these women were declared cancer-free. For others, it took years before the false alarm was confirmed. Not surprisingly, all these women suffered the anxiety, depression, family disruption, and occasional lost time from work that comes with coping with the notion that they might be suffering from a fatal cancer.17
In 1998, researchers from the University of Washington School of Medicine found that among 10,000 women screened with at least one mammogram, 1 in 4 had a false-positive test. After ten mammograms, half had had a false-positive test. These false-positive tests led to 870 outpatient visits, 539 repeat mammograms, 188 biopsies, and 1 hospitalization. The researchers estimated that among 32 million women screened for breast cancer, about 16 million will have had at least one false-positive mammogram. They concluded that “Over 10 years, one-third of women screened had abnormal test results requiring additional evaluation, even though no breast cancer was present . . . Physicians should educate women about the risk of a false positive result of a screening test for breast cancer.”18
In 2012, researchers from Oslo, Norway, reviewed European breast screening programs involving more than 390,000 women. They found that among women between ages fifty and sixty-nine who were screened for ten years, as many as 82,000 had received a false-positive diagnosis and 23,000 had had an unnecessary biopsy.19
For those trying to decide whether to undergo screening mammograms, here are the odds.
For every 1,000 women screened every two years, most will have a false-positive mammogram during the next ten years, 146 will have an unnecessary biopsy, 7 will see a fatal case of breast cancer prevented, and 19 will be diagnosed with a cancer that never would have killed them. Among the 19 per 1,000 who are overdiagnosed, 99 percent will have surgery, 70 percent radiation therapy, 70 percent hormone therapy, and 25 percent chemotherapy without benefit.20
Because the use of screening mammograms to prevent breast cancer has both risks and benefits, the recommendations from various advisory groups for when or whether to get them are all over the map:21
Women trying to decide whether to undergo annual or biennial mammograms could make one of two choices, both of which are defensible:
At the very least, women need to understand the risks and benefits of mammography. Most don’t. A recent study found that 69 percent of advertisements for cancer prevention screening programs did not quantify benefits and harms or, worse, presented them in a misleading manner.22
So, what is the answer to the question posed at the beginning of this chapter? How can we tell who will benefit from screening mammograms and who won’t? The current answer is that we can’t. Until we develop better immunological, biological, or genetic markers that distinguish breast cancers that will kill you from those that won’t, women are left with a decision that is based on incomplete information. Probably the biggest offender is ductal carcinoma in situ (DCIS), which accounts for about 20 percent of all breast cancers and might or might not progress to more invasive disease. For this reason, treatment options vary widely. At the moment, there is just no telling who is at risk and who isn’t.23
There remains no perfect choice, no risk-free strategy.
One thing, however, is clear. Once someone has made the choice for surgery, radiation, or chemotherapy, they hold on to that choice with the strength of a religious belief, which is perfectly understandable.
For example, while at dinner with an older couple, I said that I was writing a book that included, among other things, overdiagnosis and overtreatment of prostate, thyroid, and breast cancers—specifically, that many of those diagnosed with these cancers would more likely die with them than from them. Unknown to me at the time, the man had recently had a total prostatectomy and had suffered the inevitable consequence of incontinence, necessitating a pad. He was insistent, however, that his was one of the lives that had been saved. I didn’t try to dissuade him. (Had I known that he had been diagnosed with prostate cancer, I never would have brought up the subject.)
For the next few weeks, the man continued to send me emails with commentaries that he believed supported his point of view, that prostate surgery had saved his life. His response isn’t surprising. He had made the Big Bet. Once someone has undergone a thyroidectomy or a mastectomy or a prostatectomy, or has suffered through chemotherapy, radiation therapy, or hormone therapy, the only way to reduce the dissonance that they may have made the wrong choice is to believe that what they did was absolutely necessary. Otherwise, all their suffering will have been in vain.
This Big Bet phenomenon was tested years ago at the racetrack. Robert Knox and James Inkster, researchers at the University of British Columbia, went to Exhibition Park in Vancouver and asked bettors to rate their horse’s chance of winning thirty seconds before and a few seconds after making a bet on the horse to win. They titled their study “Postdecision Dissonance at Post Time.” Knox and Inkster found that bettors consistently rated their horse higher after making the bet than before making it, even though the horse, the jockey, the odds, and the previous track record hadn’t changed during the intervening minute. The only thing that had changed was that the bettors had locked in their choice.24