Navigating the Mammography Maze, Part II

PART 1:  Miscommunicating the Mammography Message.

The recent blitz of media coverage of controversial new mammography screening guidelines from an expert advisory panel has left many women confused and anxious about what they should do now and how best to navigate the maze of numbers about the risks and benefits of testing. 

As one California friend wrote:

I am fretting over the new mammogram guidelines. We all know women who were stricken in their 40s. Would love to hear your thoughts.  also fear that the timing of these recommendations is unfortunate.  Skeptics, and I guess I am one, would think about all the cost cutting that might ensue with health reform or accuse the establishment of skimping on good care... What is going on?

There is a clash between the anecdotal experiences of women like my friend Mary on one side and evidence-based medicine from long-term studies and statistical models on the other. It's easy for personal stories--and emotion--to trump science. After all, we each have a cousin, colleague or close friend whose small breast tumor was picked up in a routine mammogram.

breast cancer - Justin Sullivan.jpg
But gut instinct isn't enough when it comes to public health. And it's hard for even the best doctors to generalize from their own practices about the long-term impact of mammography on breast cancer mortality among women at various ages. It's important for women and their doctors to use this latest controversy to look closely at the studies--and statistics--used by the U.S. Preventive Services Task Force and make an informed decision about what mammography screening strategy is best for them. 

For 25 years, this independent group of experts has advised the federal government on medical and public health practices. It does not look at financial costs in its evaluations. Instead it takes a rigorous look at the scientific evidence regarding the health outcomes of mammography screening in order to come up with the experts' collective judgment about what "best practice" guidelines will provide the greatest benefits and fewest harms for the affected population of women at large. 

So what convinced the expert task force of seven women and nine men to change course, recommending that universal mass screening of women in their 40s be discontinued in favor of individualized decision making and that the time between mammograms in women 50 and over be extended to two years instead of one?

To make some sense of the numbers, I dove into 50 or so pages of scientific backup evidence published in the November 17 Annals of Internal Medicine by the task force and seven other independent research groups with whom the task force consulted. I also checked in with the author of an editorial accompanying the Preventive Services Task Force reports, Dr. Karla Kerlikowske, a University of California, San Francisco, professor of medicine, epidemiology and bio-statistics whom I have long regarded as one of the experts best able to make sense of cancer screening numbers.

Based on 30 years of covering the debate, here's my take on the most important issues to keep in mind in navigating the mammography maze:

  • SAVING A LIFE? The number that captured the most press attention is that for every  1,904 women who get screening mammograms every year in their 40s, one breast cancer death might be prevented. Not surprisingly, the most common reaction seemed to be "what if I was that one?"
But what does that "one" really mean? As with all statistics, it depends on which numbers are presented. "Relative risk" numbers are often used, but they make little sense unless you also know the underlying "absolute risk" of an event occurring.

In this case, preventing one cancer death per 1,904 women screened throughout their 40s (that's actually one death prevented for every 19,000 mammograms) translates to roughly a 15 percent relative risk reduction in breast cancer mortality, when compared with those who are not screened. But that number can be misleading. 

Because breast cancer deaths are so uncommon among women in their 40s, the task force said that the reduction is actually quite small in absolute terms. A woman's lifetime risk of dying from breast cancer is estimated to be about 2.86 percent if she doesn't start annual screening during her 40s and 2.41 percent if she does, a modest benefit of 0.45 percent, noted Kerlikowske. (These statistics apply to most women at average risk, not to those at higher risk of breast cancer.)

  • FALSE ALARMS? Suspicious mammograms that don't turn out to be cancer are relatively common in all age groups, but more so in those screened in their 40s than among those ages 50 to 74. A single mammogram screen among women ages 40 to 49 carries about a one in 10 chance of being a "false positive." Studies in community practice suggest that, cumulatively, one-quarter to one-half of women screened annually in their 40s will have at least one suspicious mammogram during the decade that later turns out to be a false alarm, said Kerlikowske.
  • MORE MEDICAL FOLLOWUP? An abnormal screening mammogram may generate a host of medical tests and procedures, from diagnostic mammograms and ultrasounds to MRIs and surgical biopsies, as well as follow-up specialty consultations. The process exacts its own toll on the affected women and their families, and the impact can vary widely. For some, the emotional impact is transient; for others, it may leave a legacy of fear and anxiety, even if they ultimately find out they don't have breast cancer. 
  • OVERDIAGNOSIS? Mammography screening has increased the detection of very small cancers. An unknown number of patients may undergo diagnosis and treatment, including surgery, chemotherapy, or radiation, for tiny, slow-growing cancers that may never have proven life-threatening in the first place.
  • READING THE IMAGES? While mammography imaging has improved dramatically over the years (and radiation exposure has been reduced), screening tests on pre-menopausal women are often harder to read than those of post-menopausal women. The breast tissue of younger women is often far more dense than that of women 50 years and older. While computers can aid in the identification of questionable areas, the interpretation of the image, and whether to label it suspicious enough to require medical follow-up, is ultimately left up to the clinical judgments of radiologists who read the films.
  • DEJA VU? Since the late 1970s, there has been controversy and uncertainty about national mammogram screening guidelines for women, particularly those in their 40s. The pendulum has swung back and forth many times, and this undoubtedly won't be the last.
The U.S. Preventive Services Task Force regularly updates its recommendations based on the evolving science. In 2002, it supported routine mammography screening every one to two years starting at age 40, with the caveat that younger women discuss the potential benefits and harms of screening mammography. Based on the latest evidence from international studies from countries like England, Sweden, and six different computer models, today's task force decided to recommend that routine mass screening begin at age 50 for women at average risk and that women in their 40s make more individualized decisions about whether to get the test.

The American College of Physicians (which represents internists and publishes the Annals of Internal Medicine) has also cautioned against routine mammography screening of most women in their 40s, as have consumer advocacy groups such as The National Breast Cancer Coalition. 

Kerlikowske noted, however, that these groups generally do not use an independent panel to make recommendations to them and do not have a published method of reviewing evidence to make guidelines, as does the Preventive Services Task Force.

  • CONSENSUS ON OLDER WOMEN? There has always been stronger scientific evidence, less uncertainty, and far more consensus about the benefits of routine screening for women 50 years and older. The risk of getting breast cancer increases with age, and the effectiveness of mammogram screening increases as well, resulting in a greater reduction in breast cancer deaths for women ages 50 to 74 than for women ages 40 to 49. The "strongest evidence for the greatest benefit is among women aged 60 to 69," the advisory panel concluded. Models also estimate that when screening is started at age 40, there are about 60 percent more false-positive results than if screening is started at age 50.
  • TIMING? Whether to do mammograms every year or every two years has been a long-running debate. The latest Preventive Services Task Force recommendations suggest that screening every two years may be effective enough to catch most moderately growing cancers that could impact a woman's survival. This is because the biology of breast cancers varies:  most breast cancers that are detected by mammography are slow to moderate growing tumors and can be detected at an early stage with biennial screening.
More aggressive deadly cancers may not be picked up early enough even with annual screens. Thus, said Kerlikowske, there is little added benefit to annual screening compared to biennial. At the same time, the two-year interval reduces the possible harms. Annual screening has almost double the number of false-positive tests compared to biennial screening.

  • OTHER HEALTH RISKS? The risk of dying from breast cancer should be put in context with other life-threatening diseases. Dartmouth Medical School researchers have shown that by the time a woman reaches age 40, heart attack, stroke, colon cancer, breast cancer, AIDS and accidents are about equally common causes of death, although death from any cause is still quite unlikely (about 17 of 1,000 women age 40 will die in the next decade of their lives). However, after age 50, heart attack becomes the single largest cause of death for women. Nonetheless, breast cancer is generally the most feared disease among women, while heart disease--the number one killer of women--is less so. Even the number one cancer killer of women--lung cancer--is less feared. 
  • INDIVIDUALIZED RISK ASSESSMENT? In the end, decision-making about when to start screening mammograms and how often to get them is really a judgment call that should be individualized to each woman's risk profile for breast cancer. Increasing age is the most important risk factor for breast cancer for most women, noted the panel, which concluded that starting breast mammography screening at age 50 would provide the biggest net benefit.
However, some women are at much higher risk because of genetic mutations that predispose them toward breast cancer or because of previous exposure to chest radiation.  Many doctors consider family history, a "first-degree relative" (parent, sibling or child), as a marker for higher risk. Women at higher risk may be advised to start mammography screening and other preventive measures at younger ages. (The National Cancer Institute lists a number of factors that may increase breast cancer risk and has an interactive breast cancer assessment tool to estimate an individual woman's chances of developing invasive breast cancer.)

Used wisely, mammography screening for breast cancer is an important tool for women's health. However, more work must be done to provide a woman and her doctor with usable risk assessment information that can clarify her own chances of getting and dying from breast cancer at various ages and how mammography screening might help (trust me, you don't want to read the Annals studies in a doctor's office!). Hopefully the U.S. Preventive Services Task Force, which did a poor job of communicating its new guidelines to the general public, will find ways to better explain the convincing risk-benefit information it has gathered.

But, in the meantime, how does an expert translate all of the complex mammography screening information into commonsense practice for her patients?  Dr. Kerlikowske, who directs the Women's Clinic at the San Francisco Veterans Administration Medical Center, explains:

In our clinic we perform mammography every two years in women aged 50-69 years and have for the last 10 years.  If a woman is in her 40s and has a first-degree relative with breast cancer, we offer her mammography or if a woman requests it. We do not routinely offer mammography to women in their 40s.
As for the doctor herself, it's simple:  "I follow the same guidelines I adhere to for my patients," she said.

Photo Credit: Getty Images Justin Sullivan