As we have suggested before in these reports, the question of whether mass, population-based screening for certain cancers (including prostate cancer) is really the best strategy for prevention of risk from the disease itself is a complex one.
A new article, published this week in BMJ, addresses this issue with specific focus on the use of mammography in mass, population-based screening for breast cancer. The full article is available on line, and we would suggest interested readers consult that full article rather than the summary media reports that have been flying around in the past 24 hours.
In this article, Jørgensen and Gøtzsche report their estimate of the extent of “overdiagnosis” (the detection of cancers that will not cause death or symptoms) in publicly organized breast cancer screening programs.
“Overdiagnosis” is not a term we like. It is loaded with the implication that the diagnosis was unnecessary. What the term really means, however, is that such a diagnosis presents a high risk of unnecessary treatment because of bias on the part of physicians and patients that every cancer — once diagnosed — must be aggressively treated. We do need to start to understand that a finding of cancer cells is not necessarily sufficient to make treatment a clinical imperative. Why? Because excessive treatment comes with serious potential for harm. In the case of breast cancer it is not unusual today for women with no family history of breast cancer and no known specific risk factors to have a double mastectomy after a finding of very early stage breast cancer (a condition called “ductal carcinoma in situ” or DCIS). In the vast majority of cases, there is no more reason for this than there would be to carry out a radical prostatectomy in a man with high-grade prostatic intraepithelial neoplasia (HG-PIN). But look at the emotional and physical cost for an unknown clinical benefit.
Anyway, the new article is a highly detailed review of findings on the diagnosis, management, and outcomes of treatment of women identified with breast cancer in five different populations (from the United Kingdom, Canada, Australia, Sweden, and Norway). It uses highly defined epidemiologic modeling systems to analyze the available data. And it comes to a very simple conclusion:
- 52 percent of all cases of breast cancer diagnosed in populations offered organized mammography screening are forms of breast cancer that will not cause either symptoms or death.
This has been translated (by the authors, not by the media) into the statement that:
- One in three cases of breast cancer is overdiagnosed.
Interestingly, this is approximately the same as the risk of “overdiagnosis” of prostate cancer currently observed in the USA.
The “New” Prostate Cancer InfoLink does not believe that, as a consequence of these data, we should start screaming that breast cancer screening is unnecessary or excessive or unwise. To the contrary, what these data imply is that mammography does not tell either a patient or her doctor whether the cancer that is being observed actually presents clinical risk. Just as with prostate cancer, it is becoming clear that we need tests that can help doctors and their patients differentiate between clinically significant and clinically insignificant forms of breast cancer based on the biologic characteristics of the cancer and the age and health of the patient.
The other thing that has to happen here is that — as a society — we need to start understanding the massive difference between a finding of some cancer cells and a diagnosis of a clinically significant cancer. We have tended, for the last 30 years, to assume that these are the same thing, and to treat every finding of cancer aggressively. It is time for us to start to use the knowledge we already have to think better about risk. In prostate cancer, active surveillance is a clear potential pathway to use the knowledge we have — and a similar strategy can be applied in the case of breast cancer too.
There are enormous emotional and psychological factors at play in all of this. It may not seem “rational” to large numbers of people not to treat some cancers aggressively. And that alone is one of the reasons why we will never overcome the problem of overtreatment without really good tests to help us separate the wheat (the truly clinically significant forms of cancer) from the chaff (the cancers that will never cause symptoms or death). The immediate and the long-term implications for cancer screening (at least in the cases of breast and prostte cancer) are vast.