Abstract
“
The Canadian National Breast Screening Study (CNBSS) [1] was initiated in 1980 and, recently, we published the results of our 25-year follow-up of all 89,835 participants, who when enrolled were aged 40–59 years [1]. They were randomly assigned with informed consent to mammography or control arms. In the mammography arm, two-view mammography was used together with breast examinations annually. In the control arm there was no mammography screening, but a single breast examination for women aged 40–49 years followed by usual care, and annual breast examinations for women aged 50–59 years. All women were taught breast self-examination (BSE) [1]. Screening was for up to 5 years. We found no evidence that mammography screening reduced breast cancer mortality. However, because of the long-term follow-up, we were able to estimate that mammography resulted in the overdiagnosis of 22% of screen-detected invasive breast cancers, 50% of the impalpable cancers detected by mammography alone. If one were to include in situ cancers, the proportion of screen-detected cancers that were overdiagnosed increases to 35%, 72% of those detected by mammography alone.
In contradistinction to the breast screening trials conducted in Sweden [2], women with breast cancer in the Canadian trial had the benefit of modern adjuvant therapy, within the Canadian universal healthcare system.
Mammography clearly has a major role in the diagnosis of breast cancer, although we have to recognize that it cannot detect all breast cancers; 31% of screen-detected breast cancers were detected by the breast examinations alone in the CNBSS in women aged 40–49 years and 19% in women aged 50–59 years [3,4]. Furthermore, women can detect breast cancers themselves in the intervals between screens, or after the last screen, either accidentally or through BSE. In the CNBSS in women aged 40–49 years, 26% of the breast cancers ascertained in the first 5 years in the mammography arm were diagnosed in this way, and 18% in women aged 50–59 years, while in the control arm the corresponding percentages were 76 and 46%, respectively [3,4]. These differences between the two arms suggest we should have found a difference in mortality from breast cancer, but that we did not shows that the early diagnosis associated with informed women with ready access to healthcare and modern therapy makes mammography screening unnecessary.
This is a conclusion that goes against all that women have been told in the last few years, as well as the conclusions of several expert groups, and has been met by incredulity, together with attempts to discredit the CNBSS and its investigators [5]. Yet, independent evaluation has confirmed our randomization was appropriate [6] resulting in a good balance of risk factors for breast cancer in the two arms [3,4]. We fully expect these met the expected indicators of good mammography [7], while our linkage to national databases ensured our follow-up was complete. Unfortunately, the ready availability of modern adjuvant therapy in the CNBSS, but not in the trials that found a benefit of screening was not considered by any of the expert groups.
We are not alone in suggesting that the risk/benefit equation for mammography screening has now shifted too far to the side of risk [8]. Advances in imaging techniques and quality are unlikely to alter this equation, as the vaunted increase in sensitivity of techniques such as digital mammography [9], or even tomosynthesis, are likely to increase the extent of overdiagnosis, without resulting in more curable breast cancers. Research is urgently needed on molecular biomarkers that will enable us to identify overdiagnosed breast cancers, to avoid their over-treatment.
“
So what are women to do now? One of the rapid responses to our report postulated that without mammography screening, developed countries would revert to the situation in some developing countries, with substantially poorer survival. That of course will not happen. The problem in most developing countries is the lack of early diagnosis of cancers, due to the lack of public information on cancer and its curability, and the lack of access to diagnosis and modern treatment. So we may need to further promote both, as they need to do even more urgently in developing countries. It is usual to claim that BSE is of no value, largely because two randomized trials failed to find a benefit [10,11], yet in one there is good evidence that the BSEs were not performed well, and in the other, evidence from the trial itself suggests the initial apparently good performance of BSE did not last long. There is good evidence from a study in Finland [12] and from within the CNBSS [13] that women who practice BSE have a reduced risk of dying from breast cancer. Extend that to some potential benefit from the breast examinations in the CNBSS [14], and a new approach begins to emerge, based upon BSE or, as some prefer to call it, increased breast awareness, and periodic examinations by a health professional trained to recognize the signs of early breast cancer as in the CNBSS, combined with prompt and free access to modern treatment for breast cancer.
The reactions to the report of the CNBSS suggest that many, perhaps particularly from the radiology community, will not be satisfied with such an approach. But as many others have demonstrated, the trends in breast cancer mortality in many countries that have embraced mammography screening are not compatible with the postulated benefits from such screening [15,16]. It is time to re-evaluate our approaches to the control of breast cancer. Prevention has a much greater role than many have recognized. It is possible to make some estimates of the proportion of breast cancer caused by various risk factors, and therefore, the amount of breast cancer potentially preventable. Poor diet and over-nutrition resulting in obesity in postmenopausal women contributes approximately 30%, as does late age at first birth, while lack of physical activity will contribute approximately 25%. Lack of breastfeeding and high alcohol consumption each contribute approximately 10%, while use of exogenous estrogens (e.g., hormone-replacement therapy) and genetics 5%. Radiation may contribute another 1%. These figures are based on data from the Western world, but are rapidly becoming applicable in some of the developing countries as well.
However, the most important approach currently has to be continuation of public and professional education for early diagnosis, combined with adequate and accessible modern therapy.
Financial & competing interests disclosure
The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.