The United States Preventivec Services Task Force (USPSTF) today released a series of reports updating their guideline recommendations for screening mammography for the early detection of breast cancer. Their conclusions are bound to raise another round of intense discussion about the benefits, risks and harms of screening for breast cancer.
There is certainly nothing wrong with that, with the exception that if we make the wrong decisions or offer women the wrong guidance about the early detection of breast cancer, we could reverse the considerable progress that has been make in reducing deaths from this disease over the past twenty years.
Unlike the Task Force, the American Cancer Society is not changing its current recommendations that women at average risk of getting breast cancer should get a mammogram every year starting at age 40.
In this era of health care reform, these new Task Force guidelines could have real implications for how insurers, government programs and maybe even the pending health care reform bills will cover screening mammography in the future.
Before I actually discuss the guidelines, I would like to set the stage with the very last sentence of the report that came from one of the evidence reports written by researchers from the Oregon University Health Sciences Center (OHSU). I do this because I think it puts the issue into context:
“Mammography screening at any age is a tradeoff of a continuum of benefits and harms. The ages at which this tradeoff becomes acceptable to individuals and society are not clearly resolved by the available evidence.” (emphasis mine)
With that as a starting point, here are the short versions of the Task Force’s new recommendations for screening mammography:
1) The Task Force recommends against routine screening for women ages 40-49. Whether to start screening before age 50 should be an individual choice.
2) The Task Force recommends screening every two years for women between ages 50 and 74.
3) The Task Force can’t make any recommendations on whether women ages 75 and over should be screened, because there is not enough evidence upon which they can base a recommendation.
4) There is not enough evidence to make a recommendation about the value of clinical breast examination (a careful breast exam done periodically by a trained medical professional) for women 40 years of age or older
5) There is no evidence that teaching women how to do breast self examination makes an difference, so they recommend against teaching women how to do it
6) There isn’t enough evidence to say anything about the value of digital mammography and MRI screening in women at average risk of breast cancer
So now the recommendations of the Society are considerably different from the Task Force, whereas in the past the only real difference was whether a screening mammogram should be done every year (ACS) or every one to two years (Task Force). Until now, both organizations had recommended starting screening for breast cancer at age 40.
Those recommendations had been in place for many years. These changes are bound to confuse women and health care professionals who must now make a professional and a personal choice as to which recommendations to follow. The worst outcome would be if the confusion leads women to do nothing since the experts can’t seem to make up their minds.
The Task Force believes their new recommendations can retain most of the benefits of mammograms—that is, decreasing deaths from breast cancer—while reducing the risks and harms of the procedure, which includes such things as having to get additional studies to clarify a suspicious finding on a mammogram, or getting a biopsy of a suspicious lesion that turns out not to be breast cancer, or perhaps having a woman embark on a treatment for an actual breast cancer that would never have interfered with her life.
The review of the various clinical trials as reported by OHSU showed that mammography reduced deaths from breast cancer by about 15% in women ages 40-49. They also found that 1904 (range 929-6378) women had to be screened over 10 years to save one life. For women ages 50-59 years, the reduction in deaths was about the same (14%). The number that needed to be screened was 1339 (range 322-7455). In women ages 60-69, the reduction in deaths was 32%, and the number who needed to be screened over 10 years was 377.
What this means is that mammograms are indeed successful in reducing deaths from breast cancer in all age groups, especially women between 60 and 69 years old. But since the actual incidence of breast cancer is less in women ages 40-49, the absolute/actual numbers of lives saved is also less. So you have to screen more women to get the same benefit.
Stated another way, the Task Force agrees that mammography reduces deaths in women ages 40-49. It just doesn’t save enough lives, in their opinion.
What about those risks and harms of getting a mammogram? Here is what did the OHSU investigators have to say:
- No significant damage was seen from the radiation associated with mammograms.
- Mammograms can be painful, but “few (women) would consider this a deterrent from future screening.”
- There was no consistent effect on most women with regards to the anxiety associated with mammograms, but it was an issue for some women.
- “False positive” mammograms—where the screening mammogram suggests there may be a cancer, but eventually none is found—are an issue, with more of them in younger women compared to older women. But false positive mammograms that lead to an actual biopsy are less common in younger women than in older women, which means that younger women may need more extra mammograms or ultrasounds to take a look at a suspicious area but don’t actually have to have a biopsy done when compared to older women where the opposite is true. (In more precise terms, according to the paper, in women ages 40-49, for every case of invasive breast cancer that is diagnosed 556 women have a mammogram, 47 have additional images, and 5 have biopsies.)
- Overdiagnosis was a difficult issue to address, because there really is no direct way of determining which breast cancers we treat are cancers that might lead to a woman’s death as compared to breast cancers we treat that would never cause a problem. They concluded that overdiagnosis rates in various studies ranged from 1% to 30%, with most falling between 1% to 10%.
As the Oregon researchers point out based on this analysis, “These estimates are difficult to apply because, for individual women, it is not known which types of cancer will progress, how quickly cancer will advance and expected lifetimes.”
The largest burden of overdiagnosis probably occurs in the population of older women, where you can diagnose and treat a breast cancer but woman wouldn’t have a problem with the breast cancer because she had another serious disease and died from something other than breast cancer. If that is where the bulk of the problem lies, then that is a different situation than having overdiagnosis in a young woman, where it could impact the quality of her life for many more years.
What about new technologies such as digital mammograms (which are quickly becoming the only type of mammogram available in many cities in this country) and MRI screening for women at average risk of breast cancer?
Here is what the OHSU researchers to say about those topics as well as a comment about how often mammograms should be done:
“New technologies, such as digital mammography and MRI, have become widely used in the United States without definitive studies of their effect on screening. Consumer expectations that new technology is better than old may obscure potential adverse effects, such as higher false-positive results and expense. No screening trials incorporating newer technology have been published, and estimates of benefits and harms in this report are based predominantly on studies of film mammography. No definitive studies of the appropriate interval for mammography screening exist, although trial data reflect screening intervals from 12 to 33 months.”
Let’s now focus on the other research report which was based on a very sophisticated computer model designed and supported by the National Cancer Institute. The purpose of this model was to try and determine at what age screening mammography should begin, when it should end, and how often it should be done.
The model actually looked at 20 different age/frequency “scenarios.” Six different institutions around the country that participate in this project looked at each of these scenarios and came up with their own estimates of how the different combinations of age and frequency impacted the benefits of getting a screening mammogram.
I suspect to no one’s surprise, each of these six complex computer models came up with different answers for the same questions.
For example, in one model, if you screened only women from 50-74 and did it every two years, you reduced breast cancer deaths by about 28%. If you did it every year from age 40 to 84, you reduced mortality by about 54%. In another model, the same numbers were about 22% and 38%. In the first study, doing mammograms every other year for more years made a big difference. In the second study, it still made a difference, but not quite as much. And there were still other studies where it made little or no difference
And, not unexpectedly, the later you started getting a screening mammogram and how often you did it resulted in a significant difference in the number of mammograms a woman would have over her lifetime. Start later, end earlier and get it every two years required many fewer lifetime mammograms than starting at 40, screening to a later age, and getting it every year.
So what did these experts conclude from their computer models?
“This study uses 6 established models that use common inputs but different approaches and assumptions to extend previous randomized mammography screening trial results to the US population and to age groups in whom trial results are less conclusive. All 6 modeling groups concluded that the most efficient screening strategies are those that include a biennial screening interval. Conclusions about the optimal starting ages for screening depend more on the measure chosen for evaluating outcomes. If the goal of a national screening program is to reduce mortality in the most efficient manner, then programs that screen biennially from age 50 years to age 69, 74 or 79 years are among the most efficient on the basis of the ratio of benefits to the number of screening examinations. If the goal of a screening program is to efficiently maximize the number of life-years gained, then the preferred strategy would be to screen biennially starting at age 40 years. Decisions about the best starting and stopping ages also depend on tolerance for false-positive results and the rate of overdiagnosis.”
The bottom line of this research was that you could get somewhere between 70-99% of the benefit of screening mammograms (that is, reducing deaths from breast cancer) while reducing the harms by about 50% if you started screening at age 50 and did it every two years as compared to starting at age 40 and doing it every year.
Eventually, someone has to take this information and make some recommendations, and that is exactly what the Task Force did.
We probably have learned as much as we are going to learn from large clinical trials of mammography. If we are going to extend our knowledge about the benefits, risks and harms of mammography, it probably won’t come from new, large clinical trials. We have to find other ways to answer our questions about the early detection of breast cancer, and one of the ways to do that is through computer models.
The question, however, is whether or not the models are sufficiently accurate to tell us with reasonable certainty what would happen under a particular situation. It is one thing to try to predict the future or support a theory. It is quite a different thing, in my opinion, when you take computer models and make public policy that affects millions of women with respect to a life threatening disease. Even though the models may be very well designed, there are always questions about how well they truly reflect or predict “real life.”
Aside from the confusion this report is going to sow in the minds of women about when (and maybe even whether) they should be screened for breast cancer, there is the question about how we are going to provide insurance coverage for women who need mammograms.
It remains to be seen how insurers, Medicare, Medicaid and states where insurers are required to cover screening mammograms are going to react to these recommendations. Hopefully, they will continue to recognize that other respected organizations—such as the American Cancer Society—have different thoughts on this issue and are still appropriate benchmarks to use when determining whether or not to pay for screening mammograms.
And then there is health care reform, where the influence of the Task Force may be considerable under the various legislative proposals currently wending their way through Congress.
If the Task Force recommendations become the benchmark in the new legislation, then we may have a problem. If that turns out to be the case, hopefully Congress will realize that recommendations from other organizations that have looked at the same evidence and who have come to different conclusions should also be considered as valid when making coverage decisions for new or existing insurance plans. If not, then it will be much more difficult for a woman to get a mammogram if she is between 40 and 49 years old, or if she wants to get one every year as we currently recommend.
The American Cancer Society is not changing our recommendations for breast cancer screening as a result of this report. Based on our initial review of this new guideline, we see no reason to change a strategy that has proven effective in reducing the death rates for breast cancer in all age groups, including those women ages 40-49.
We will review the evidence offered by the computer modeling approach since it represents new research, and we will continue to examine information from other sources as it becomes available. And, if that information or research is compelling, we will always be open to updating our recommendations. But until such time as we are convinced that such evidence supports such a change, our guidelines will remain as they have been for the past 12 years.
What we know—as noted in the Task Force report—is that deaths from breast cancer have declined 2.3% per year for all women and 3.3% per year for women aged 40-50 years beginning in 1990. That may not seem much year to year, but the total impact over 19 years has been significant, and cannot be ignored. This is especially true when one considers that the death rate was absolutely stable for the preceding six decades. Screening mammograms and better treatments are responsible for that success.
We do no agree that 70% of the benefit from screening mammograms is the right way to go. We do believe that we should aim to choose 100% of the benefit. We should not forget that the “benefit” in this situation is reducing deaths from breast cancer. A 30% reduction in saving lives is not acceptable.
We also recognize that mammograms are not perfect. We realize that women do have to get additional studies for suspicious lesions. We realize that some women have biopsies that do not show breast cancer. We realize that our predictive tests are not perfect, so that we can’t say with certainty which breast cancers are aggressive and require intensive treatment and which would—if left alone—never cause a problem.
We realize that we need better screening tools, and that we must work diligently to improve the quality of screening mammography across the country.
Until we have something better, what we have to work with to detect breast cancer early is the screening mammogram. Is it imperfect? Yes. Has it saved lives and reduced deaths from breast cancer? Absolutely.
And that is the fact that simply cannot be ignored.